doc_id
string | appl_id
string | claim1
string | claim_number
int64 | claim_text
string |
|---|---|---|---|---|
8205155
|
12529071
|
1. <nc>A method</nc> of substituting <nc>alternative text</nc> in <nc>an electronic input</nc> of <nc>an author's work</nc> to <nc>a text management system</nc>, comprising: storing <nc>text fragments</nc> of <nc>previous text</nc>; indexing <nc>the stored text fragments</nc> on <nc>at least two different bases</nc>; allowing <nc>the author</nc> to enter <nc>text</nc>; comparing <nc>text fragments</nc> in <nc>the entered text</nc> to <nc>the stored text fragments</nc>, where one of <nc>the text fragments</nc> in <nc>the entered text</nc> is substantially similar on one of <nc>the at least two different bases</nc> to one of <nc>the stored text fragments</nc>; displaying to <nc>the author</nc> <nc>a substitute text fragment</nc> for <nc>the one text fragment</nc> in <nc>the entered text</nc>; and providing <nc>the author</nc> with <nc>the option</nc> of substituting <nc>the substitute text fragment</nc> for <nc>the one text fragment</nc> of <nc>the entered text</nc>, wherein a first one of <nc>the at least two different bases</nc> is solely <nc>an edit distance basis</nc> and a second one of <nc>the at least two different bases</nc> combines <nc>a word occurrence algorithm</nc>, <nc>a word difference algorithm</nc>, and <nc>an edit distance algorithm</nc>, and <nc>the two bases</nc> contribute differently to <nc>a final similarity basis</nc>.
| 3
|
3. <nc>The method</nc> of substituting <nc>alternative text</nc> as claimed in <nc>claim</nc> 1 , wherein when storing <nc>text fragments</nc> of <nc>previous text, related words</nc> are resolvable to <nc>a single normalized word form</nc>.
|
9298679
|
13790285
|
1. <nc>A computer-implemented method</nc> for providing <nc>a binary representation</nc> of <nc>a web page</nc>, <nc>the method</nc> comprising: parsing <nc>a web page source document</nc>, using <nc>a processor</nc>, to identify <nc>one or more page elements</nc>, <nc>the source document</nc> comprising <nc>text</nc> <nc>that</nc> defines <nc>a web page</nc>, where <nc>the source document</nc> is written in <nc>hypertext markup language</nc> (<nc>“HTML</nc>”); generating, with <nc>the processor</nc>, <nc>a binary representation</nc> corresponding to <nc>a document object model structure</nc> of <nc>a web page</nc> using <nc>the identified one or more page elements</nc>, <nc>the binary representation</nc> including <nc>a conversion</nc> of <nc>a type</nc> of <nc>each page element</nc>, <nc>syntax information</nc> of <nc>each page element</nc>, and <nc>the one or more page elements</nc> to <nc>a format</nc> other than <nc>plain text</nc>, wherein <nc>the binary representation</nc> is generated using <nc>a binary representation dictionary</nc>, <nc>the binary representation dictionary</nc> defining <nc>at least one binary instruction</nc> corresponding to <nc>each</nc> of <nc>the one or more page elements</nc>; and providing <nc>the binary representation</nc> such that, in <nc>response</nc> to <nc>a request</nc> from <nc>a client device</nc>, <nc>the binary representation</nc> is provided to <nc>the client device</nc> to render <nc>the web page</nc> without parsing <nc>the web page source document</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the binary representation</nc> includes <nc>a datum</nc> identifying <nc>a version</nc> of <nc>the binary representation dictionary</nc>.
|
7869989
|
11343251
|
1. <nc>A method</nc> for representing <nc>text</nc> in <nc>a language independent format</nc> of <nc>multiple bits</nc> and <nc>fields</nc> of <nc>bits</nc> comprising: referencing <nc>a word</nc> from <nc>a set</nc> of <nc>words</nc> into <nc>a dictionary</nc> of <nc>representations</nc> of <nc>words</nc>, in <nc>said format</nc> of <nc>multiple bits</nc> and <nc>fields</nc> of <nc>bits</nc>, to find <nc>a corresponding dictionary term</nc>, <nc>the set</nc> of <nc>words</nc> derived from <nc>a natural language usage</nc> defining <nc>a natural language context</nc> of <nc>the words</nc> in <nc>the set</nc> of <nc>words</nc>; mapping <nc>the referenced dictionary term</nc> to <nc>at least one definition element</nc> indicative of <nc>usage</nc> of <nc>the word</nc> in <nc>at least one context</nc>, <nc>the definition elements</nc> defining <nc>a record</nc> of <nc>fields</nc>; comparing <nc>the mapped definition element</nc> to <nc>the corresponding fields</nc> in <nc>the definition elements</nc> of <nc>other words</nc> in <nc>the set</nc> of <nc>words</nc>, <nc>the comparison operative</nc> to identify <nc>similar contexts</nc> between <nc>the definition elements</nc>; disambiguating <nc>the referenced words</nc> by analyzing <nc>each</nc> of <nc>the definition elements</nc> of <nc>the referenced words</nc> with <nc>the definition elements</nc> of <nc>the other referenced words</nc>, <nc>the analysis</nc> operable to determine <nc>a particular definition</nc> for <nc>each</nc> of <nc>the referenced words</nc> in <nc>the context</nc> of <nc>the set</nc> of <nc>words</nc>, disambiguating including: performing, with <nc>a processor</nc>, bitwise operations on <nc>at least a subset</nc> of <nc>the fields</nc> in <nc>the definition elements</nc> with <nc>corresponding fields</nc> in <nc>the other definition elements</nc> in <nc>the set</nc> of <nc>words</nc>, <nc>each</nc> of <nc>the definition elements</nc> indicative of <nc>a particular context</nc>, <nc>the operations</nc> for identifying <nc>a particular definition element</nc> based on <nc>the context</nc> in <nc>the set</nc> of <nc>words</nc>, <nc>the definition elements</nc> including <nc>bit fields</nc> of <nc>class</nc>, <nc>method</nc>, and <nc>category</nc> in <nc>the high bits</nc>, <nc>the method field</nc> determining <nc>a structure</nc> of <nc>fields</nc> in <nc>lower order bits</nc>, wherein <nc>disambiguating</nc> is performed on <nc>definition elements</nc> of <nc>equal category</nc> and <nc>method fields</nc>; and identifying, from <nc>the comparing</nc>, <nc>a definition element</nc> corresponding to <nc>the usage</nc> of <nc>the word</nc> in <nc>a context</nc> of <nc>the set</nc> of <nc>words</nc>.
| 16
|
16. <nc>The method</nc> of <nc>claim</nc> 1 further comprising <nc>a plurality</nc> of <nc>definition elements</nc>, <nc>each word</nc> having <nc>one definition element</nc> with <nc>a primary definition element value</nc> and <nc>at least one definition element</nc> with <nc>a multiple taxonomy value</nc>, <nc>the multiple taxonomy value</nc> indicative of <nc>different contexts</nc> of <nc>usage</nc> for <nc>the word</nc>, <nc>the primary definition element</nc> denoted by <nc>a predetermined value</nc> in <nc>a field</nc> of <nc>the definition element</nc>.
|
9733821
|
14195325
|
1. <nc>A method</nc>, comprising: at <nc>an electronic device</nc> comprising <nc>a processor</nc> and <nc>memory</nc> storing <nc>instructions</nc> for <nc>execution</nc> by <nc>the processor</nc>: while <nc>the device</nc> is operating with <nc>a first setting</nc> in <nc>a first state</nc>, <nc>detecting</nc>, at <nc>a first time</nc>, a change in <nc>settings</nc> of <nc>the device</nc> to change <nc>the first setting</nc> from <nc>the first state</nc> to <nc>a second state</nc> <nc>that</nc> is different from <nc>the first state</nc>; while <nc>the device</nc> is operating with <nc>the first setting</nc> in <nc>the second state</nc>, receiving, at <nc>a second time</nc> <nc>that</nc> is after <nc>the first time</nc>, a user input <nc>that</nc> corresponds to <nc>a pattern</nc> of <nc>user behavior</nc>, wherein <nc>the user input</nc> is <nc>a user voice input</nc>; in <nc>response</nc> to receiving <nc>the user input</nc>: comparing <nc>the pattern</nc> of <nc>user behavior</nc> to <nc>a plurality</nc> of <nc>predefined conditions</nc> <nc>that</nc>, when met, indicate that <nc>the user</nc> is having <nc>difficulty</nc> with operating <nc>the device</nc>, wherein <nc>a predefined condition</nc> of <nc>the plurality</nc> of <nc>predefined conditions</nc> includes <nc>the user voice input</nc> containing <nc>one or more predetermined words</nc> associated with <nc>user difficulty</nc>; in <nc>accordance</nc> with <nc>a determination</nc>, based on <nc>the comparison</nc> of <nc>the pattern</nc> of <nc>user behavior</nc> to <nc>the plurality</nc> of <nc>predefined conditions</nc>, that <nc>the device</nc> changed <nc>the first setting</nc> from <nc>the first state</nc> to <nc>the second state</nc> within <nc>a predetermined time period</nc> prior to receiving <nc>the user input</nc>, restoring <nc>the first setting</nc> to <nc>the first state</nc>; and in <nc>accordance</nc> with <nc>a determination</nc>, based on <nc>the comparison</nc> of <nc>the pattern</nc> of <nc>user behavior</nc> to <nc>the plurality</nc> of <nc>predefined conditions</nc>, that <nc>the user</nc> is not having <nc>difficulty</nc> with operating <nc>the device</nc>, maintaining <nc>the first setting</nc> in <nc>the second state</nc>.
| 11
|
11. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: prior to restoring <nc>the first setting</nc> to <nc>the first state</nc>, prompting <nc>the user</nc> to confirm that <nc>they</nc> want to restore <nc>the first setting</nc> to <nc>the first state</nc>; and in <nc>response</nc> to receiving <nc>confirmation</nc> that <nc>the user</nc> wants to restore <nc>the first setting</nc> to <nc>the first state</nc>, restoring <nc>the first setting</nc> to <nc>the first state</nc>.
|
8027806
|
12453651
|
1. <nc>A computer-implemented method</nc> of constructing <nc>a test</nc>, comprising: receiving at <nc>a computer</nc> <nc>a selection</nc> for <nc>a test item</nc> for <nc>inclusion</nc> in <nc>a set</nc> of <nc>test items</nc>; updating with <nc>the computer</nc> <nc>a correlation value</nc> based on <nc>the set</nc> of <nc>test items</nc> including <nc>the selected test item</nc>, wherein <nc>the correlation value</nc> indicates <nc>an extent</nc> of <nc>correlation</nc> between <nc>the set</nc> of <nc>test items</nc> and <nc>a test specification</nc>, <nc>the test specification</nc> comprising one or more of <nc>an overall test difficulty rating</nc>, <nc>a correlation</nc> between <nc>a correct response</nc> for <nc>a selected test item</nc> and <nc>a particular cognitive or behavioral trait</nc>, <nc>an orientation</nc> of <nc>the presentation</nc> of <nc>questions</nc> and <nc>answers</nc> for <nc>the set</nc> of <nc>selected test items</nc>, <nc>a number</nc> of <nc>pages</nc> of <nc>text</nc> for <nc>a test</nc>, <nc>a mean point-biserial</nc>, <nc>a mean r-biserial</nc>, and <nc>an arrangement</nc> of <nc>the set</nc> of <nc>selected test items</nc>; determining with <nc>the computer</nc> whether <nc>the set</nc> of <nc>test items</nc> including <nc>the selected item</nc> satisfies <nc>the test specification</nc>; if <nc>the set</nc> of <nc>test items</nc> including <nc>the selected test item</nc> does not satisfy <nc>the test specification</nc>, receiving <nc>input</nc> at <nc>the computer</nc> revising <nc>the set</nc> of <nc>test items</nc> in <nc>order</nc> to satisfy <nc>the test specification</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , <nc>the test item</nc> comprising <nc>a question</nc> and <nc>one or more answers</nc>.
|
9437195
|
14030919
|
1. <nc>A system</nc> comprising: <nc>a user speech profile</nc> stored on <nc>a computer readable storage device</nc>, <nc>the speech profile</nc> containing <nc>a plurality</nc> of <nc>phonemes</nc> with <nc>user</nc> identifying <nc>characteristics</nc> for <nc>the phonemes</nc>, wherein the identifying characteristic comprise <nc>a representation</nc> of <nc>the amount</nc> of <nc>pronunciation difference</nc> for <nc>each</nc> of <nc>the plurality</nc> of <nc>phonemes</nc> from <nc>an average user</nc>; and <nc>a speech processor</nc> coupled to access <nc>the speech profile</nc> and generate <nc>a phrase</nc> containing user distinguishing <nc>phonemes</nc> based on <nc>the pronunciation difference</nc> between <nc>the user</nc> identifying <nc>characteristics</nc> for <nc>such phonemes</nc> and <nc>average user</nc> identifying <nc>characteristics</nc>, such that <nc>the phrase</nc> has <nc>discriminability</nc> from <nc>other users</nc> wherein <nc>the speech processor</nc> generates <nc>the phrase</nc> by searching for <nc>a phrase</nc> in <nc>a library</nc> <nc>that</nc> contains <nc>user distinguishing phonemes</nc>, wherein <nc>a user distinguishing phoneme</nc> is based on <nc>pronunciation</nc> of <nc>a phoneme</nc> by <nc>a user</nc> differing from <nc>the way</nc> <nc>most other users</nc> pronounce <nc>the phoneme</nc> and wherein <nc>such distinguishing phonemes</nc> have <nc>an associated score</nc> indicative of <nc>a magnitude</nc> by <nc>which</nc> <nc>such pronunciation</nc> differs and wherein <nc>the associated score</nc> is adjusted based on <nc>a difference</nc> of <nc>the pronunciation</nc> from <nc>ambient noise</nc>, resulting in <nc>a different generated phrase</nc> in <nc>the presence</nc> of <nc>the ambient noise</nc>.
| 4
|
4. <nc>The system</nc> of <nc>claim</nc> 1 wherein <nc>the speech processor</nc> further receives <nc>the generated phrase</nc> spoken by <nc>the user</nc> via <nc>the audio input</nc> and verifies whether <nc>the phrase</nc> was spoken by <nc>the user</nc> to confirm <nc>the identity</nc> of <nc>the user</nc> and allow <nc>access</nc> to <nc>the system</nc>.
|
8700544
|
13162906
|
1. <nc>A method</nc> comprising: receiving <nc>a query</nc> q from <nc>a particular user u</nc> <nc>who</nc> intends to find <nc>results</nc> <nc>that</nc> satisfy <nc>the query</nc> q with <nc>respect</nc> to <nc>a topic</nc> <nc>T u</nc> , <nc>the particular user</nc> <nc>u</nc> being characterized by <nc>user information</nc> <nc>θ u</nc> ; producing <nc>a generic topic distribution</nc> <nc>Pr r</nc> (<nc>T|q</nc>) associated with the query q <nc>that</nc> is germane to <nc>a population</nc> of <nc>generic users</nc>; producing <nc>a user-specific query-dependent topic distribution</nc> <nc>Pr(T u |q</nc>,<nc>θ u</nc> ) associated with <nc>the query q</nc> for <nc>the particular user u</nc>; generating <nc>personalized results</nc> for <nc>the particular user u</nc> based on <nc>the generic topic distribution</nc> Pr r (<nc>T|q</nc>) and the user-specific query-dependent topic distribution Pr(T u |q,<nc>θ u</nc> ); and forwarding <nc>the personalized results</nc> to <nc>the particular user</nc> <nc>u.</nc>
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising, in <nc>a training process</nc>, learning <nc>a user-specific query-independent distribution</nc> <nc>Pr(T u |θ u</nc> ), <nc>the training process</nc> comprising: generating <nc>training data</nc> associated with <nc>the particular user u</nc>, <nc>the training data</nc> identifying <nc>queries</nc> <nc>q t</nc> submitted by the particular user u and N distributions of <nc>topics</nc> (T) t associated with <nc>search results</nc>, <nc>the search results</nc> being provided by <nc>search engine functionality</nc> in <nc>response</nc> to <nc>the queries</nc> <nc>q t</nc> ; and computing <nc>the user-specific query-independent distribution</nc> <nc>Pr(T u |θ u</nc> ) by computing <nc>an average</nc> over <nc>the N distributions</nc> of <nc>topics</nc> <nc>(T) t</nc> .
|
8555234
|
12553224
|
1. <nc>A method</nc> for <nc>verification</nc> of <nc>soft error resilience</nc> of <nc>devices</nc> from <nc>design data</nc> for <nc>a logic design</nc>, comprising: selecting <nc>a statistically relevant set</nc> of <nc>Soft Error Rate (SER) sensitive logic devices</nc> deemed critical to <nc>a SER robustness</nc> for <nc>a test process</nc> of <nc>the logic design</nc> by using <nc>a SER categorization tool</nc> to generate a SER sensitive device test set, and further comprising: successively filtering <nc>the design data</nc> through <nc>a series</nc> of <nc>heuristic rule-based device identifiers</nc> that group and annotate <nc>the SER sensitive logic devices</nc> by parsing <nc>a netlist</nc> representing <nc>the logic design</nc> for <nc>identification nomenclature</nc>, <nc>device type</nc> and <nc>connectivity</nc>; selecting, <nc>grouping</nc>, and associating <nc>related devices</nc> for <nc>testing</nc> according to <nc>a list</nc> of <nc>predetermined SER sensitive device types</nc>; and reducing <nc>a number</nc> of <nc>the devices</nc> in <nc>the SER sensitive device test</nc> set based on <nc>symmetry</nc> of <nc>structures</nc> joined to <nc>a common bus</nc> and further reducing <nc>flow device vectors</nc> to <nc>scalars</nc>; and inducing <nc>a logic state fault</nc> in <nc>successive ones</nc> of <nc>said reduced number</nc> of <nc>devices</nc> in <nc>said test process</nc> of <nc>the logic design</nc> for <nc>functional design verification</nc> <nc>which</nc> provides <nc>a sufficient delay</nc> after <nc>injection</nc> for <nc>the logic design</nc> to detect <nc>a logic fault</nc> for <nc>those reduced number</nc> of <nc>devices</nc> in <nc>said test process</nc>; and providing <nc>logic fault information</nc> of <nc>said detected logic fault</nc> of <nc>said reduced number</nc> of <nc>devices</nc> to facilitate <nc>corrective action</nc>.
| 6
|
6. <nc>The method</nc> according <nc>to claim</nc> 1 further comprising: performing <nc>heuristic examination</nc> based on <nc>the identification nomenclature</nc> and <nc>the connectivity</nc> to extract and store device lists, <nc>the device</nc> lists including one or more of: <nc>a logic flow list</nc>, <nc>a finite state machine list</nc>, <nc>a control register list</nc>, and <nc>an untested device list</nc>.
|
8201096
|
11760759
|
1. <nc>A computer-implemented method</nc>, comprising: performing, at <nc>a computing device</nc>, <nc>a search</nc> using <nc>a search query</nc>, wherein <nc>the search</nc> is through <nc>indexed content</nc> and <nc>metadata</nc> associated with <nc>a plurality</nc> of <nc>files</nc> in <nc>a hierarchical file system</nc>; generating <nc>search results</nc>; determining <nc>an associated file type</nc> for <nc>each file</nc> in <nc>the search results</nc>; using <nc>the file type</nc> to identify <nc>a plug</nc>-in capable of processing <nc>the content</nc> in <nc>the file</nc> associated with <nc>that file type</nc>; using <nc>the plug-in</nc> for <nc>each file</nc> to process <nc>the content</nc> for <nc>that file</nc> and dynamically generate <nc>a preview view entry</nc> for <nc>that file</nc>, wherein <nc>the preview view entry</nc> includes <nc>a display</nc> of <nc>the content</nc> of <nc>the file</nc> or <nc>an icon</nc> representing <nc>the content</nc> of <nc>the file</nc>; generating <nc>a preview view</nc> including <nc>one or more preview view entries</nc> in <nc>a preview view area</nc>; generating <nc>a list view</nc> including <nc>one or more list view entries</nc> in <nc>a list view area</nc>; linking <nc>the list view</nc> and <nc>the preview</nc> view together such that scrolling or sorting within <nc>either view area</nc> concurrently causes <nc>the same scrolling</nc> or sorting within <nc>the other view area</nc>; and concurrently displaying <nc>the same search results</nc> in <nc>both view areas</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the list view area</nc> and <nc>preview area</nc> are <nc>separate and distinct areas</nc>, and wherein when <nc>a preview view entry</nc> is interactive, <nc>the preview view entry</nc> includes <nc>one or more interactive controls</nc>.
|
7692685
|
11094922
|
1. <nc>An object tracker system</nc>, comprising: <nc>a processor</nc> <nc>that</nc> executes <nc>the following computer executable components</nc> stored on <nc>a computer readable medium</nc>: <nc>an audio model component</nc> <nc>that</nc> models <nc>an original audio signal</nc> of <nc>an object</nc>, a time delay between <nc>at least two audio input signals</nc> and <nc>a variability component</nc> of <nc>the original audio signal</nc>, <nc>the audio model</nc> employing <nc>a probabilistic generative model</nc>; <nc>a video model component</nc> <nc>that</nc> models <nc>a location</nc> of <nc>the object</nc>, <nc>an original image</nc> of <nc>the object</nc> and <nc>a variability component</nc> of <nc>the original image</nc>, <nc>the video model</nc> employing <nc>a probabilistic generative model</nc>, <nc>the video model</nc> receiving <nc>a video input</nc>; and <nc>an audio video tracker component</nc> <nc>that</nc> models <nc>the location</nc> of <nc>the object</nc> based, at least in <nc>part</nc>, upon <nc>the audio model</nc> and <nc>the video model</nc>, wherein <nc>the audio video tracker</nc> provides <nc>an output</nc> associated with <nc>the location</nc> of <nc>the object</nc> based on, at least in <nc>past</nc>, <nc>a linear mapping</nc> <nc>that</nc> approximates <nc>the location</nc> of <nc>the object</nc>, wherein <nc>the linear mapping</nc> is computed as <nc>a function</nc> of <nc>the time delay</nc> between <nc>the at least two audio input signals</nc>, <nc>wherein error</nc> in approximating <nc>the location</nc> of <nc>the object</nc> is modeled by <nc>a zero mean Gaussian distribution</nc> associated with <nc>a precision matrix</nc>, and wherein <nc>the zero mean Gaussian distribution</nc> associated with <nc>the precision matrix</nc> is based on, at least in <nc>part</nc>: <nc>a product</nc> of <nc>a horizontal position</nc> of <nc>the object</nc> and <nc>a difference</nc> in <nc>horizontal position</nc> of <nc>a first audio input device</nc> and <nc>a second audio input device</nc>; <nc>a product</nc> of <nc>a vertical position</nc> of <nc>the object</nc> and <nc>a difference</nc> in <nc>vertical position</nc> of <nc>the first audio input device</nc> and <nc>the second audio input device</nc>: and <nc>a precision matrix</nc> of <nc>an approximation error</nc> modeled by <nc>a zero mean Gaussian</nc>.
| 6
|
6. <nc>The system</nc> of <nc>claim</nc> 1 , <nc>wherein the original audio signal</nc> of <nc>the object</nc>, <nc>the time delay</nc> between <nc>at least two audio input signals</nc>, and <nc>the variability component</nc> of <nc>the original audio signal</nc> comprise <nc>unobserved variables</nc> of <nc>the audio model</nc>; and wherein <nc>the audio model</nc> further includes <nc>trainable parameters</nc>.
|
9858264
|
15466102
|
1. <nc>A method</nc> for converting <nc>a text sentence</nc> into <nc>an image sentence</nc>, <nc>the method</nc> comprising: receiving <nc>a text sentence</nc> comprising <nc>a plurality</nc> of <nc>words</nc>, wherein <nc>the text sentence</nc> includes <nc>a verb phrase</nc>, and wherein <nc>the text sentence</nc> is associated with <nc>a plurality</nc> of <nc>semantic roles</nc>; querying <nc>an image database</nc>, <nc>the image database</nc> having stored therein <nc>a plurality</nc> of <nc>candidate images</nc>, <nc>at least a portion</nc> of <nc>which</nc> are tagged with <nc>image content descriptors</nc>; making <nc>a determination</nc> that <nc>none</nc> of <nc>the candidate images</nc> in <nc>the image database</nc> captures <nc>each</nc> of <nc>the plurality</nc> of <nc>semantic roles</nc> associated with <nc>the text sentence</nc>; in <nc>response</nc> to making <nc>the determination</nc>, generating <nc>first and second sentence fragments</nc> from <nc>the text sentence</nc>, wherein <nc>each</nc> of <nc>the first and second sentence fragments</nc> is associated with <nc>a respective first and second fragmented semantic role</nc>; identifying <nc>a first subset</nc> of <nc>the candidate images</nc> <nc>that</nc> are stored in <nc>the image database</nc>, <nc>each</nc> of <nc>which</nc> captures <nc>the first fragmented semantic role</nc>; generating <nc>a first feature vector</nc> characterizing <nc>the first subset</nc> of <nc>candidate images</nc> stored in <nc>the image database</nc>; and identifying a first particular one of <nc>the candidate images</nc> in <nc>the first subset</nc> <nc>that</nc> is characterized by <nc>features</nc> closest to <nc>a first mean value</nc> derived from <nc>the first feature vector</nc>.
| 9
|
9. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: identifying <nc>a first word</nc> in <nc>the received text sentence</nc> as being <nc>a verb</nc>; identifying <nc>a second word</nc> in <nc>the received text sentence</nc> as being <nc>an adjunct</nc>; and identifying <nc>the verb</nc> <nc>phrase</nc> as comprising <nc>the first and second words</nc>, wherein <nc>the first word</nc> is included in <nc>the first sentence fragment</nc>, and <nc>the second word</nc> is included in <nc>the second sentence fragment</nc>.
|
8526735
|
13461856
|
1. <nc>A method</nc> for processing <nc>a time-series analysis</nc> of <nc>keywords</nc>, <nc>the method</nc> comprising: segmenting, with <nc>a processor</nc>, by performing at least one of <nc>clustering</nc> and <nc>classifying</nc>, <nc>pieces</nc> of <nc>document data</nc> based at least in <nc>part</nc> on <nc>frequencies</nc> of <nc>occurrence</nc> of <nc>keywords</nc> in <nc>the pieces</nc> of <nc>document data</nc>, wherein <nc>the pieces</nc> of <nc>document data</nc> include <nc>a description</nc> in <nc>a natural language</nc>, <nc>the segmenting</nc> resulting in creating <nc>at least one document cluster</nc> and <nc>at least one keyword cluster</nc>; and acquiring <nc>a frequency distribution</nc> showing <nc>variation</nc> of <nc>the frequencies</nc> of <nc>occurrence</nc> of <nc>the pieces</nc> of <nc>document data</nc> by performing, with <nc>the processor</nc>, at least one of: <nc>a time-series analysis</nc> of <nc>frequencies</nc> of <nc>occurrence</nc> of <nc>pieces</nc> of <nc>document data</nc> containing <nc>individual keywords</nc> in <nc>at least one document cluster</nc>, and <nc>a time-series analysis</nc> of <nc>frequencies</nc> of <nc>occurrence</nc> of <nc>pieces</nc> of <nc>document data</nc> containing <nc>at least one keyword cluster</nc>.
| 4
|
4. <nc>The method</nc> according to <nc>claim</nc> 1 , further comprising: receiving <nc>a search query</nc>; narrowing down <nc>a set</nc> of <nc>pieces</nc> of <nc>document data</nc> by <nc>the search query</nc> to create <nc>a narrowed set</nc> of <nc>pieces</nc> of <nc>document data</nc>; approximating the narrowed set with <nc>a linear sum</nc> of <nc>document clusters</nc>; and <nc>inferring</nc>, from <nc>at least one frequency distribution</nc> of <nc>a keyword</nc> for <nc>each document cluster</nc>, <nc>at least one frequency distribution</nc> of <nc>the keyword</nc> in <nc>the narrowed set</nc> of <nc>pieces</nc> of <nc>document data</nc>.
|
9922639
|
13739826
|
1. <nc>A system</nc>, comprising: <nc>a speech interface platform</nc> configured to (a) receive <nc>an audio signal</nc> from <nc>a speech interface device</nc> located in <nc>an environment</nc> of <nc>a user</nc>, (b) recognize <nc>user speech</nc> in <nc>the audio signal</nc>, (c) understand <nc>a user intent</nc> from <nc>the recognized user speech</nc>, and (d) perform <nc>an action</nc> based on <nc>the understood user intent</nc>; <nc>a graphical interface</nc> <nc>that</nc> is provided by, and accessible by <nc>the user</nc> through, <nc>a user device</nc> <nc>that</nc> is physically separate from <nc>the speech interface device</nc>, <nc>the graphical interface</nc> being configured to display <nc>a first interface</nc> including <nc>a list</nc> of <nc>a plurality</nc> of <nc>historical interaction records</nc>, <nc>a historical interaction record</nc> of <nc>the plurality</nc> of <nc>historical interaction records</nc> associated with <nc>information</nc> regarding <nc>a historical speech interaction</nc> between <nc>the speech interface platform</nc> and <nc>the user</nc>, wherein <nc>the information</nc> includes <nc>(a) the recognized user speech</nc>, (b) <nc>the understood user intent</nc>, and (c) <nc>the action</nc> performed by <nc>the speech interface platform</nc>; wherein <nc>the first interface</nc> provided by, and accessible by <nc>the user</nc> through, <nc>the user device</nc> <nc>that</nc> is physically separate from <nc>the speech interface device</nc> is further configured to accept, within <nc>the first interface</nc> provided by <nc>the user device</nc>, a selection of <nc>the historical interaction record</nc>; wherein <nc>the graphical interface</nc> is configured to display, based at least in <nc>part</nc> on <nc>the selection</nc>, <nc>a second interface</nc>, <nc>the second interface</nc> configured to display <nc>at least a portion</nc> of <nc>the information</nc> regarding <nc>the historical speech interaction</nc> and to accept, within <nc>the second interface</nc> provided by <nc>the user device</nc>, <nc>one or more edits</nc> or <nc>evaluations</nc> from <nc>the user</nc> of <nc>the historical speech interaction</nc> regarding one or more of (a) <nc>the understood user intent</nc> and (b) the action performed by <nc>the speech interface platform</nc>; and wherein <nc>the system</nc> is configured to receive, from <nc>the user device</nc> <nc>that</nc> is physically separate from <nc>the speech interface device</nc>, <nc>the one or more edits</nc> or <nc>evaluations</nc> for improving <nc>future speech recognition</nc> and intent <nc>understanding</nc> on <nc>a subsequent audio signal</nc> received from <nc>the speech interface device</nc> located in <nc>the environment</nc> of <nc>the user</nc>.
| 29
|
29. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the list</nc> of <nc>the plurality</nc> of <nc>historical interaction records</nc> is presented in <nc>reverse chronological order</nc>.
|
9317189
|
14136288
|
1. <nc>A method</nc>, comprising: receiving, via <nc>an element</nc> included in <nc>a user interface</nc>, <nc>a user-defined, structured input</nc>; selecting, in <nc>real time</nc>, <nc>an active input</nc> from <nc>the user-defined, structured input</nc> based at least in <nc>part</nc> on <nc>context</nc> associated with <nc>the user interface</nc>; using <nc>a processor</nc> to determine <nc>a context-sensitive rule</nc> <nc>that</nc> applies to <nc>the active input</nc>, wherein: <nc>the context-sensitive rule</nc> requires that <nc>the active input</nc> be <nc>a specific data type</nc>; and using <nc>the processor</nc> to determine <nc>the context-sensitive rule</nc> includes: setting one or more of <nc>the following</nc>: <nc>a permitted values</nc> bit or a permitted formats bit; dividing, based at least in <nc>part</nc> on <nc>the active input, inactive inputs</nc> from <nc>the user-defined, structured input</nc> into <nc>two groups</nc>: <nc>(1) those inactive inputs</nc> <nc>that</nc> affect <nc>the active input</nc> with <nc>respect</nc> to <nc>a format or content limitation</nc> and <nc>(2) those inactive inputs</nc> <nc>that</nc> do not affect <nc>the active input</nc> with <nc>respect</nc> to <nc>a format or content limitation</nc>; determining if <nc>any inactive inputs</nc> affect <nc>the active input</nc> with <nc>respect</nc> to <nc>a format</nc> or <nc>content limitation</nc>; and in <nc>the event</nc> <nc>it</nc> is determined <nc>none</nc> of <nc>the inactive inputs</nc> affect <nc>the active input</nc> with <nc>respect</nc> to <nc>a format or content limitation</nc>: determining, without taking into <nc>consideration</nc> <nc>any</nc> of <nc>the inactive inputs</nc>, if <nc>the active input</nc> is limited to <nc>certain permitted values</nc>; and determining, without taking into <nc>consideration</nc> <nc>any</nc> of <nc>the inactive inputs</nc>, if <nc>the active input</nc> is limited to <nc>certain permitted formats</nc>; and providing, in <nc>real time</nc> via <nc>the user interface</nc>, <nc>guidance</nc> associated with <nc>the active input</nc> and <nc>the context-sensitive rule</nc>, including by performing the following: displaying, in <nc>the user interface</nc>, <nc>format assistance</nc> <nc>which</nc> includes <nc>one or more of the following: (1) identification</nc> of <nc>the specific data type</nc> <nc>that</nc> satisfies <nc>the context-sensitive rule</nc> or <nc>(2) automatic configuration</nc> of <nc>the user interface</nc> so that <nc>the active input</nc> has <nc>a data type</nc> <nc>which</nc> matches <nc>the specific data type</nc> <nc>that</nc> satisfies <nc>the context-sensitive rule</nc>; and displaying, in <nc>the user interface</nc>, <nc>format validation</nc> <nc>which</nc> indicates whether <nc>the context-sensitive rule</nc> is satisfied based at least in <nc>part</nc> on <nc>a current data type</nc> of <nc>the active input</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein providing <nc>guidance</nc> includes using one or more of <nc>the following</nc>: <nc>a still image</nc>, <nc>an audio file</nc>, <nc>a video file</nc>, <nc>a CSS style</nc>, <nc>a hyperlink</nc>, <nc>a tooltip</nc>, <nc>a help box</nc>, <nc>a dropdown box</nc>, <nc>automatic configuration</nc> of <nc>the user interface</nc>, <nc>a highlight</nc>, or <nc>an error message</nc>.
|
9691031
|
15193526
|
1. A method programmed in <nc>a non-transitory memory</nc> of <nc>a device</nc> comprising: <nc>a.</nc> automatically analyzing <nc>target information</nc>; <nc>b.</nc> automatically fact checking <nc>the target information</nc> by comparing <nc>the target information</nc> with <nc>source information</nc> to generate <nc>a result</nc>, wherein comparing includes <nc>at least one of: i.</nc> searching for <nc>an exact match</nc> of <nc>the target information</nc> in <nc>the source information</nc> and returning <nc>the exact match search result</nc> of <nc>the exact match search</nc> if <nc>the exact match</nc> is found; <nc>ii</nc>. utilizing pattern matching for <nc>fact</nc> checking and returning <nc>the result</nc> of the pattern matching fact check if <nc>a pattern matching result confidence score</nc> is above <nc>a pattern matching result confidence threshold</nc>; and <nc>iii</nc>. utilizing <nc>a natural language search</nc> for <nc>fact</nc> checking and returning <nc>the result</nc> of <nc>the natural language fact</nc> check if a natural language result <nc>confidence score</nc> is above <nc>a natural language result confidence threshold</nc>; and <nc>c.</nc> automatically presenting <nc>a status</nc> of <nc>the target information</nc> in <nc>real-time</nc> based on <nc>the result</nc> of <nc>the comparison</nc> of <nc>the target information</nc> with <nc>the source information</nc>, wherein searching for <nc>the exact match</nc> begins searching <nc>the source information</nc> controlled by <nc>a media company</nc>, then using <nc>crowdsourced data</nc> as <nc>the source information</nc>, and then using <nc>world wide web data</nc> for <nc>fact</nc> <nc>checking</nc>; wherein utilizing <nc>pattern matching</nc> begins utilizing <nc>the source information</nc> controlled by <nc>the media company</nc>, then using <nc>the crowdsourced data</nc> as <nc>the source information</nc>, and then using <nc>the world wide web data</nc> for <nc>fact</nc> checking; and wherein <nc>the natural language search</nc> begins searching <nc>the source information</nc> controlled by <nc>the media company</nc>, then using <nc>the crowdsourced data</nc> as <nc>the source information</nc>, and then using <nc>the world wide web data</nc> for <nc>fact</nc> <nc>checking</nc>, wherein searching for <nc>the exact match</nc> begins searching <nc>the source information</nc> located in <nc>a designated fact checking database</nc>, then goes to <nc>a broader set</nc> of <nc>source information</nc>, and repeatedly goes to <nc>broader sets</nc> of <nc>source information</nc> until <nc>a broadest source information</nc> set has been exhausted; wherein utilizing <nc>pattern matching</nc> begins utilizing <nc>the source information</nc> located in <nc>the designated fact checking database</nc>, then goes to <nc>the broader set</nc> of <nc>source information</nc>, and repeatedly goes to <nc>broader sets</nc> of <nc>source information</nc> until <nc>the broadest source information</nc> set has been exhausted; and wherein <nc>the natural language search</nc> begins searching <nc>the source information</nc> located in <nc>the designated fact checking database</nc>, then goes to <nc>the broader set</nc> of <nc>source information</nc>, and repeatedly goes to <nc>broader sets</nc> of <nc>source information</nc> until <nc>the broadest source information</nc> set has been exhausted.
| 12
|
12. <nc>The method</nc> of <nc>claim</nc> 1 further comprising utilizing <nc>a plurality</nc> of <nc>fact</nc> checking <nc>implementations</nc> initially, wherein <nc>each fact</nc> checking <nc>implementation</nc> utilizes <nc>a different set</nc> of <nc>source information</nc>, and comparing <nc>results</nc> of <nc>each fact</nc> checking <nc>implementation</nc>, and iteratively eliminating <nc>a fact</nc> checking <nc>implementation</nc> of <nc>the plurality</nc> of <nc>fact</nc> checking <nc>implementations</nc> with <nc>a lowest confidence score</nc> until <nc>a single fact</nc> checking implementation remains.
|
7769810
|
11796037
|
1. <nc>A computer-implemented system</nc> comprising: <nc>a memory</nc>; and <nc>at least one processor</nc> coupled to <nc>the memory</nc>, the at least one processor to implement: <nc>a viewer</nc> to open <nc>a master copy</nc> of <nc>an electronic document</nc> in <nc>a local editor</nc> for <nc>display</nc> at <nc>a display device</nc>; <nc>a first queue</nc> associated with <nc>the local editor</nc>, <nc>the first queue</nc> being to store <nc>edit operations</nc> requested by <nc>the local editor</nc>, <nc>the first queue</nc> maintained by <nc>the local editor</nc>; <nc>a second queue</nc> associated with <nc>a remote editor</nc>, <nc>the remote editor</nc> residing at <nc>a remote client computer</nc>, <nc>the second queue</nc> being to store <nc>edit operations</nc> requested by <nc>the remote editor</nc>, <nc>the second queue</nc> maintained by <nc>the local editor</nc>; <nc>an update detector</nc> to detect <nc>a network request</nc> from <nc>the remote editor</nc> to perform <nc>an edit operation</nc> on <nc>a remote copy</nc> of <nc>the electronic document</nc>, <nc>the remote copy</nc> of <nc>the electronic document</nc> being opened by <nc>the remote editor</nc>; <nc>an update module</nc> to: in <nc>response</nc> to <nc>the network request</nc> from <nc>the remote editor</nc> to perform <nc>the edit operation</nc> on <nc>the remote copy</nc> of <nc>the electronic document</nc>, perform <nc>the edit operation</nc> on <nc>the master copy</nc> of <nc>the electronic document</nc>; and update <nc>the second queue</nc> with <nc>the edit operation</nc> in <nc>response</nc> to <nc>the performing</nc> of <nc>the edit operation</nc> on <nc>the master copy</nc> of <nc>the electronic document</nc>; and a distributor to propagate, via <nc>a network communication</nc>, <nc>the edit operation</nc> to <nc>the remote copy</nc> of <nc>the electronic document</nc>, <nc>the edit operation</nc> requested to be performed on <nc>the remote copy</nc> of <nc>the electronic document</nc> and performed on <nc>the master copy</nc> of <nc>the electronic document</nc>; wherein <nc>the viewer</nc>, <nc>the first queue</nc>, <nc>the second queue</nc>, <nc>the update detector</nc>, <nc>the update module</nc>, <nc>the local editor</nc> and <nc>the distributor</nc> are provided at <nc>a local client computer</nc>.
| 9
|
9. <nc>The system</nc> of <nc>claim</nc> 1 comprising <nc>a storing module</nc> to: detect, at <nc>the local editor</nc>, that <nc>the master copy</nc> of <nc>the electronic document</nc> is <nc>the only copy</nc> of <nc>the electronic document</nc> being opened; detect <nc>a request</nc> to close <nc>the master copy</nc> of <nc>the electronic document</nc>; and store <nc>the latest version</nc> of <nc>the electronic document</nc>, <nc>the latest version</nc> of <nc>the electronic document</nc> including <nc>results</nc> of <nc>operations</nc> stored in <nc>the local queue</nc> and <nc>the remote queue</nc>.
|
7610185
|
11957281
|
1. <nc>A system</nc> for navigating <nc>categorized information</nc>, <nc>the system</nc> comprising: <nc>a processor</nc>; <nc>a memory</nc> coupled <nc>the processor</nc>; <nc>computer code</nc> loaded into <nc>the memory</nc> for performing <nc>the functions</nc> of: displaying <nc>a two-dimensional map</nc> to <nc>a user</nc>, <nc>the map</nc> showing <nc>search terms</nc> relating to <nc>a subject matter</nc>, where <nc>a position</nc> of <nc>each search term</nc> corresponds to <nc>relationships</nc> between <nc>all the search terms</nc>; dynamically changing <nc>a display</nc> of <nc>the terms</nc> and <nc>a position</nc> of <nc>the search terms</nc> relative to each other based on <nc>user input</nc> into <nc>the map</nc>, <nc>the display</nc> corresponding to <nc>relative importance</nc> of <nc>the search terms</nc> to <nc>the subject matter</nc>; changing <nc>a context</nc> of <nc>a search</nc> and selecting <nc>a different branch</nc> of <nc>a taxonomic tree</nc> in <nc>response</nc> to <nc>a user input</nc> to select one of <nc>the search terms</nc> to display <nc>different search results</nc> based on <nc>a combination</nc> of <nc>different branches</nc>; displayed separately from <nc>the map</nc>, <nc>a plurality</nc> of <nc>hyperlinks</nc> corresponding to <nc>the different search results</nc>; <nc>a neural network</nc> underlying <nc>the map</nc>, wherein <nc>the display</nc>, <nc>the position</nc> of <nc>the search terms</nc> on <nc>the map</nc> and <nc>a selection</nc> of <nc>the displayed search terms</nc> are derived from <nc>the neural network</nc>, and displaying to <nc>the user</nc> <nc>additional search terms</nc> derived from <nc>the search results</nc> by <nc>the neural network</nc>.
| 3
|
3. <nc>The system</nc> of <nc>claim</nc> 1 , <nc>wherein positioning</nc> of <nc>a cursor</nc> over one of <nc>the search terms</nc> rearranges <nc>the search terms</nc> on <nc>the map</nc> to correspond to <nc>an increased relevance</nc> of <nc>that search term</nc>, while <nc>the cursor</nc> is over <nc>that search term</nc>.
|
9953088
|
15063223
|
1. <nc>A method</nc> for providing <nc>a response</nc> to <nc>a user request</nc>, comprising: at <nc>an electronic device</nc> with <nc>one or more processors</nc> and <nc>memory</nc>: receiving <nc>a user request</nc>, <nc>the user request</nc> including <nc>at least a speech input</nc> and seeks <nc>an informational answer</nc> or <nc>performance</nc> of <nc>a task</nc>, wherein: <nc>the user request</nc> is associated with <nc>a detected failure</nc> to provide <nc>a satisfactory response</nc> to <nc>the user request</nc>; and <nc>one or more crowd</nc> sourcing <nc>information sources</nc> relevant to <nc>the user request</nc> are queried in <nc>response</nc> to <nc>the detected failure</nc> to provide <nc>a satisfactory response</nc> to <nc>the user request</nc>; and generating <nc>a response</nc> to <nc>the user request</nc> based on <nc>the one or more answers</nc> obtained from querying <nc>the one or more crowd</nc> sourcing <nc>information sources</nc>.
| 12
|
12. <nc>The method</nc> of <nc>claim</nc> 1 , wherein querying <nc>the one or more crowd</nc> sourcing <nc>information sources</nc> includes querying <nc>one or more remote sources</nc>.
|
8849761
|
13615999
|
1. <nc>A method</nc> for <nc>scheduling storage operations</nc> on <nc>a cloud storage site</nc>, comprising: receiving <nc>multiple new requests</nc> for <nc>cloud storage</nc> from <nc>one or more clients</nc>, <nc>wherein the multiple new requests</nc> <nc>each</nc> include <nc>a request</nc> for <nc>data storage</nc>, and wherein <nc>the multiple new requests</nc> <nc>each</nc> include <nc>information</nc> associated with <nc>the data storage</nc> requested; determining <nc>a current capacity</nc> of <nc>the cloud storage site</nc>, wherein <nc>the current capacity</nc> of <nc>the cloud storage site</nc> is determined based at least in <nc>part</nc> on: <nc>(i</nc>) <nc>a capacity policy</nc>, wherein <nc>the capacity policy</nc> specifies <nc>preferences</nc> and <nc>criteria</nc> associated with allocating <nc>system resources</nc> for <nc>the cloud storage site</nc>, and <nc>(ii</nc>) at least one of: <nc>a quotation policy</nc>, wherein <nc>the quotation policy</nc> includes <nc>a set</nc> of <nc>preferences</nc> and <nc>criteria</nc> associated with generating <nc>a quote</nc> in <nc>response</nc> to <nc>received client requests</nc>, and <nc>a scheduled job</nc>, wherein <nc>the scheduled job</nc> is associated with <nc>a quote</nc> for <nc>cloud storage</nc> accepted by <nc>a client</nc>, <nc>a quoted job</nc>, wherein <nc>the quoted job</nc> is associated with <nc>a quote</nc> for <nc>cloud storage</nc> provided to <nc>a client</nc>, and <nc>queued requests</nc>, wherein <nc>queued requests</nc> include <nc>requests</nc> by <nc>clients</nc> for <nc>cloud storage</nc> for <nc>which</nc> <nc>the respective clients</nc> have not been provided <nc>a quote</nc>; identifying <nc>one or more approved requests</nc>, wherein <nc>the one or more approved requests</nc> are identified from <nc>pending requests</nc> based at least in <nc>part</nc> on <nc>preferences</nc> and <nc>criteria</nc> specified in <nc>the accessed quotation policy</nc> and <nc>the current capacity</nc>, wherein <nc>pending requests</nc> comprise <nc>the received multiple new requests</nc> and <nc>queued requests</nc>; generating <nc>a responsive quote</nc> for <nc>each approved request</nc>, wherein <nc>the responsive quotes</nc> are generated based at least in <nc>part</nc> on <nc>preferences</nc> and <nc>criteria</nc> specified in <nc>the accessed quotation policy</nc>, and wherein <nc>each responsive quote</nc> includes <nc>one or more pricing values</nc>; sending <nc>a generated responsive quote</nc> to <nc>a client</nc> associated with <nc>an approved request</nc> that <nc>the responsive quote</nc> was generated for; and receiving from <nc>the client</nc> <nc>that</nc> was sent <nc>the generated responsive quote</nc> <nc>an indication</nc> of <nc>acceptance</nc> of <nc>the generated responsive quote</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein, <nc>the capacity policy</nc> specifies <nc>system resources</nc> available during <nc>specified periods</nc>, <nc>scheduled maintenance windows</nc> and <nc>current storage capacity</nc> available on <nc>servers</nc>.
|
8036415
|
11619454
|
1. <nc>A method</nc> for <nc>nano-encoding information</nc> related to <nc>printed texts</nc> or <nc>images</nc> on <nc>paper</nc> and <nc>other surfaces</nc>, <nc>the method</nc> comprising: fabricating <nc>composite material</nc> by mixing <nc>nano-particles</nc>, <nc>the composite material</nc> comprising <nc>a signature</nc>; collocating <nc>the fabricated composite nano</nc><nc>-particles</nc> with <nc>the printed text</nc> or <nc>images</nc>; and <nc>nano</nc>-decoding <nc>the information</nc>, wherein <nc>nano</nc>-decoding <nc>the information</nc> comprises: detecting <nc>nano particles</nc>; determining <nc>invariant properties</nc> of <nc>the detected nano particles</nc>; matching <nc>invariant properties</nc> with <nc>predetermined coded information</nc>; analyzing <nc>the invariant properties</nc> of <nc>the detected nano particles</nc> for <nc>segmentation</nc>, in <nc>response</nc> to detecting that there is <nc>no match</nc>, and redetecting <nc>the nano particles</nc> in <nc>response</nc> to determining that <nc>the nano particles</nc> exhibit <nc>segmentation</nc>; and providing <nc>user</nc> with <nc>decoded information</nc>.
| 2
|
2. <nc>The method</nc> as in <nc>claim</nc> 1 wherein detecting <nc>nano particles</nc> comprises detecting <nc>luminescent nano particles</nc>.
|
9355089
|
14564068
|
1. <nc>A method</nc> for <nc>intention detection</nc> in <nc>information</nc>, <nc>the method</nc> comprising: parsing, using <nc>a processor</nc> and <nc>a memory</nc>, <nc>a new information</nc> into <nc>a constituent set</nc> of <nc>complete grammatical constructs</nc>, <nc>the new information</nc> being in <nc>a language</nc> and relating to <nc>a subject matter domain</nc>; identifying, in <nc>a subset</nc> of <nc>the complete grammatical constructs</nc>, <nc>a set</nc> of <nc>linguistic styles</nc> of <nc>the language</nc> according to <nc>a subset</nc> of <nc>a set</nc> of <nc>word-style associations</nc> related to <nc>the language</nc> and independent of <nc>the subject matter domain</nc>; assigning <nc>a first weight</nc> to <nc>a first linguistic style</nc> and <nc>a second weight</nc> to <nc>a second linguistic style</nc>, <nc>the set</nc> of <nc>linguistic styles</nc> including <nc>the first linguistic style</nc> and <nc>the second linguistic style</nc>; constructing, using <nc>second set</nc> of <nc>linguistic styles</nc> of <nc>the language</nc>, and using <nc>a taxonomy</nc> of <nc>intentions</nc> present in <nc>a training information</nc> relating to <nc>the subject matter domain</nc>, <nc>a set</nc> of <nc>style-intention rules</nc>, <nc>the set</nc> of <nc>style-intention rules</nc> including <nc>the style-intention rule</nc>, <nc>the style-intention rule</nc> relating <nc>a selected intention</nc> from <nc>the taxonomy</nc> of <nc>intentions</nc> with <nc>a selected linguistic style</nc> in <nc>the set</nc> of <nc>linguistic styles</nc> to indicate that <nc>the selected intention</nc> is present where <nc>the selected linguistic style</nc> appears in <nc>any information</nc> related to <nc>the subject matter domain</nc>; mapping <nc>a first intention information</nc> to <nc>the first style</nc> using <nc>a first style-intention rule</nc>, and <nc>a second intention information</nc> to <nc>the second style</nc> using <nc>a second style-intention rule</nc>; and tagging <nc>a complete grammatical construct</nc> in <nc>the subset</nc> with <nc>the first intention information</nc> responsive to <nc>a weight</nc> associated with <nc>the first intention information</nc> exceeding <nc>an intention selection threshold</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: constructing, using <nc>a lexicon</nc> of <nc>the language</nc>, <nc>the set</nc> of <nc>word-style associations</nc>, <nc>wherein a word</nc> in <nc>a word-style association</nc> is used in <nc>the language</nc> to convey <nc>the corresponding style</nc> in <nc>the word-style association</nc>.
|
8983805
|
13325885
|
1. <nc>A computer-implemented method</nc> for updating <nc>a 3D model</nc>, <nc>the method</nc> comprising: by <nc>a computer device</nc> with <nc>a processor</nc>, <nc>associated memory</nc>, and <nc>a graphical user interface</nc>, <nc>the processor</nc>: storing in <nc>the associated memory</nc>, <nc>a 3D model</nc>, said storing includes storing <nc>a history</nc> of <nc>the 3D model</nc> having undergone <nc>one or more operations</nc> and <nc>graphs</nc> linked to <nc>different points</nc> of <nc>the history</nc>; retrieving from <nc>the associated memory</nc>, <nc>at least two graphs</nc> of <nc>the stored graphs</nc> linked to <nc>the different points</nc> of <nc>the history</nc>, including: <nc>an old input graph</nc> of <nc>the 3D model</nc> before undergoing <nc>an operation</nc> of <nc>the one or more operations</nc>, and an old output graph of <nc>the 3D model</nc> corresponding to <nc>an application</nc> of <nc>the operation</nc> to <nc>the old input graph</nc>; generating <nc>a new input graph</nc> of <nc>the operation</nc>, wherein <nc>the new input graph</nc> is generated as <nc>a result</nc> of <nc>a modification</nc> of <nc>the 3D model</nc>, <nc>the modification</nc> being requested by <nc>a user</nc> via <nc>the graphical user interface</nc>; computing <nc>a double push-out rewriting rule</nc> <nc>that</nc> specifies: (<nc>i</nc>) <nc>a part</nc> of <nc>the old input graph</nc> to be replaced, (<nc>ii</nc>) <nc>a part</nc> of <nc>the new input graph</nc> to replace <nc>the part</nc> of <nc>the old input graph</nc>, and (iii) <nc>an interface</nc> <nc>which</nc> is <nc>a part</nc> common to <nc>the part</nc> of <nc>the old input graph</nc>, <nc>the part</nc> of <nc>the new input graph</nc>, and the old output graph, <nc>the rewriting rule</nc> corresponding to <nc>logical operations</nc> <nc>which</nc> are computed based on <nc>the old input graph</nc>, <nc>the old output graph</nc>, and the new input graph; and applying <nc>the rewriting rule</nc> directly on <nc>the old output graph</nc>, such that <nc>the old output graph</nc> is transformed to <nc>a new output graph</nc> <nc>that</nc> represents <nc>an updated state</nc> of <nc>the 3D model</nc> as updated by <nc>the modification</nc>.
| 10
|
10. <nc>The method</nc> of <nc>claim</nc> 1 , <nc>wherein the steps</nc> of retrieving <nc>graphs</nc>, including <nc>the old input graph</nc> and <nc>the old output graph</nc>, generating <nc>the new input graph</nc>, and computing <nc>the rewriting rule</nc> are iterated over <nc>historical operations</nc> of <nc>the 3D model</nc>, wherein <nc>the operation</nc> of <nc>each iteration</nc> being <nc>a respective historical operation</nc>.
|
9137220
|
14055276
|
1. <nc>A method</nc> for collaboratively editing <nc>a document</nc> in <nc>a system</nc> of <nc>sharee clients</nc>, <nc>the method</nc> comprising: creating <nc>a document change</nc>; generating <nc>a document</nc> token for encrypting <nc>the document change</nc>; encrypting, using <nc>a processor</nc>, <nc>the document change</nc> with the document token, wherein <nc>the encrypted document change</nc> is loaded onto <nc>one or more Cloud servers</nc>; making <nc>the encrypted document change</nc> available to <nc>sharee clients</nc>; forming <nc>a share document</nc> token by splitting <nc>the document</nc> token into a server document token and a sharee document token; generating <nc>a plurality</nc> of <nc>copies</nc> of <nc>the sharee document</nc> token; encrypting <nc>each sharee document</nc> token with <nc>a respective sharee's public key</nc>; distributing <nc>each encrypted sharee document</nc> token to <nc>respective sharee clients</nc>; wherein <nc>each sharee client</nc> is configured to: decrypt the encrypted sharee document token using <nc>a respective private key</nc>; combine <nc>the decrypted sharee document</nc> token with the server document token; decrypt <nc>the encrypted document change</nc> using <nc>the combined share document token</nc>; and consolidate <nc>the decrypted document change</nc> into <nc>the document</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the encrypted document</nc> is stored in <nc>a document server</nc>.
|
6081665
|
08994393
|
1. <nc>A real-time virtual machine method</nc> (<nc>RTVMM</nc>) for implementing <nc>real-time systems</nc> and <nc>activities</nc>, <nc>the RTVMM</nc> comprising <nc>the steps</nc>: implementing <nc>an O-OPL program</nc> <nc>that</nc> can run on <nc>computer systems</nc> of <nc>different designs</nc>, <nc>an O-OPL program</nc> being based on <nc>an object-oriented programming language</nc> (<nc>O-OPL</nc>) comprising <nc>object type declarations</nc> called <nc>classes</nc>, <nc>each class definition</nc> describing <nc>the variables</nc> <nc>that</nc> are associated with <nc>each object</nc> of <nc>the corresponding class</nc> and <nc>all</nc> of <nc>the operations</nc> called <nc>methods</nc> <nc>that</nc> can be applied to <nc>instantiated objects</nc> of <nc>the specified type</nc>, <nc>a "method</nc>" being <nc>a term</nc> of <nc>art</nc> describing <nc>the unit</nc> of <nc>procedural abstraction</nc> in <nc>an object-oriented programming system</nc>, <nc>an O-OPL program</nc> comprising <nc>one or more threads</nc> wherein <nc>the run-time stack</nc> for <nc>each thread</nc> is organized so as to allow <nc>accurate identification</nc> of <nc>type-tagged pointers</nc> contained on <nc>the stack</nc> without requiring <nc>type tag information</nc> to be updated each time <nc>the stack's content changes</nc>, <nc>the O-OPL</nc> being <nc>an extension</nc> of <nc>a high-level language</nc> (<nc>HLL</nc>) exemplified by <nc>Java</nc>, <nc>HLL</nc> being <nc>an extension</nc> of <nc>a low-level language</nc> (<nc>LLL</nc>) exemplified by <nc>C</nc> and <nc>C++</nc>, <nc>a thread</nc> being <nc>a term</nc> of <nc>art</nc> for <nc>an independently-executing task</nc>, <nc>an O-OPL program</nc> being represented at <nc>run time</nc> by <nc>either O-OPL byte codes</nc> or by <nc>native machine codes</nc>.
| 11
|
11. <nc>The RTVMM</nc> of <nc>claim</nc> 1 wherein <nc>the implementing step</nc> comprises <nc>the step</nc>: causing <nc>an application thread</nc> <nc>that</nc> is to be preempted to provide <nc>notification</nc> as to when <nc>the thread</nc> is at <nc>a point</nc> where <nc>safe garbage collection</nc> can take <nc>place</nc>.
|
8311835
|
10652685
|
1. <nc>A client-server system</nc> for providing <nc>assisted multi-modal dialogue</nc>, comprising: <nc>a web server</nc> for generating <nc>client-side markups</nc> having <nc>recognition</nc> and <nc>audible prompting</nc> for <nc>execution</nc> on <nc>a client</nc> having <nc>recognition capabilities</nc>, <nc>the web server</nc> further including <nc>controls</nc> for generating <nc>the client side markups</nc>, <nc>the controls</nc> including <nc>speech controls inheritance</nc> for setting <nc>values</nc> to <nc>properties</nc> associated with <nc>the controls</nc> and organized in <nc>collections</nc> to construct <nc>a dialog</nc> for obtaining <nc>information</nc> pertaining to <nc>a plurality</nc> of <nc>topics</nc>, <nc>each collection</nc> of <nc>controls</nc> configured to create <nc>a separate dialog</nc> associated with <nc>a separate topic</nc>; <nc>a recognition server</nc>, coupled to <nc>the web server</nc>, for providing <nc>speech recognition processing</nc> to <nc>received voice data</nc> based on <nc>a grammar or language model</nc> provided with <nc>the received voice data</nc> to produce <nc>speech recognition results</nc>, <nc>the speech recognition results</nc> being provided to <nc>the web server</nc>; and <nc>a telephone voice browser</nc>, coupled to <nc>the web server</nc>, for <nc>processing voice data</nc>, <nc>the telephone voice browser</nc> including <nc>a media server</nc> for providing <nc>a telephony interface</nc> and <nc>a voice browser</nc>; wherein <nc>the controls</nc> of <nc>the web server</nc> include <nc>companion controls</nc> associated with <nc>corresponding primary controls</nc> for providing <nc>recognition</nc> and <nc>audible prompting</nc>, the companion controls including <nc>a semantic map</nc>, wherein <nc>the semantic map</nc> includes <nc>semantic items</nc> and forms <nc>an association</nc> between <nc>a visual domain</nc> of <nc>the primary controls</nc> and <nc>a non-visual recognition domain</nc> of <nc>the companion controls</nc> and wherein <nc>the dialog</nc> includes <nc>at least one question</nc> provided by <nc>a prompt object</nc> and <nc>at least one answer</nc>, <nc>a grammar object</nc> is provided to define <nc>a grammar</nc> for <nc>recognition</nc> of <nc>input data</nc> and <nc>related processing</nc> on <nc>the input</nc> and <nc>an answer property associates</nc> <nc>a recognized result</nc> with <nc>a semantic item</nc> in <nc>the semantic map</nc>.
| 4
|
4. <nc>The client-server system</nc> of <nc>claim</nc> 1 , wherein <nc>the web server</nc> includes <nc>an authoring tool</nc> for dynamically generating <nc>the client-side markups</nc> and <nc>a specific form</nc> of <nc>markup</nc> for <nc>the type</nc> of <nc>client</nc> accessing <nc>the web server</nc>, <nc>the web server</nc> further including <nc>a library</nc> for providing <nc>visual, recognition and audible prompting markup information</nc>.
|
8249875
|
13072009
|
1. <nc>A method</nc> for analyzing <nc>a voice</nc> of <nc>a speaker</nc>, comprising: receiving, by <nc>a computer</nc>, <nc>data</nc> indicative of <nc>speech</nc> from <nc>the speaker</nc>; storing, by <nc>a computer</nc>, <nc>the received data</nc> in <nc>at least one database</nc>; calculating, by <nc>a computer</nc>, based upon <nc>the received data</nc>, an average intensity value for <nc>each</nc> of <nc>a plurality</nc> of <nc>frequencies</nc>, wherein <nc>the calculation</nc> of <nc>the average intensity value</nc> for <nc>each frequency</nc> is based on: <nc>i) dividing</nc>, by <nc>a computer</nc>, <nc>the received data</nc> into <nc>a number</nc> of <nc>time periods</nc>; <nc>ii</nc>) obtaining, by <nc>a computer</nc>, <nc>an intensity</nc> of <nc>the speaker's speech</nc> for <nc>each frequency</nc> during <nc>each time period</nc>; iii) obtaining, by <nc>a computer</nc>, a sum of <nc>intensity values</nc> for <nc>each frequency</nc> during <nc>all time periods</nc>; and iv) dividing, by <nc>a computer</nc>, the sum of <nc>intensity values</nc> for <nc>each frequency</nc> by <nc>the number</nc> of <nc>time periods</nc>; calculating, by <nc>a computer</nc>, based upon <nc>the received data</nc>, <nc>a maximum intensity value</nc> for <nc>each</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>, wherein <nc>the maximum intensity value</nc> for <nc>each frequency</nc> is <nc>the highest intensity</nc> of <nc>the speaker's speech</nc> for <nc>each frequency</nc> during <nc>all time periods</nc>; calculating, by <nc>a computer</nc>, a level of <nc>a survival element</nc> of <nc>a personality profile</nc> of <nc>the speaker</nc> based upon at least one of: (a) <nc>a rapid change</nc> in <nc>the average intensity value</nc> between <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>; and (b) a rapid change in <nc>the maximum intensity value</nc> between <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>; calculating, by <nc>a computer</nc>, a level of <nc>a homeostasis element</nc> of <nc>the personality profile</nc> of <nc>the speaker</nc> by measuring <nc>a distance</nc> between <nc>the average intensity value</nc> and <nc>the maximum intensity</nc> for <nc>each frequency</nc> of <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>; calculating, by <nc>a computer</nc>, <nc>a level</nc> of <nc>a growth element</nc> of <nc>the personality profile</nc> of <nc>the speaker</nc> based upon at least one of: <nc>(a) determining</nc>, by <nc>a computer</nc>, a frequency range within <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc> in <nc>which</nc> <nc>the average intensity value</nc> of <nc>each frequency</nc> within <nc>the frequency range</nc> is higher than <nc>a value</nc> <nc>that</nc> is equal to <nc>a predetermined percent</nc> of <nc>the highest average intensity value</nc> within <nc>the frequency range</nc> of <nc>the at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>, (<nc>b) determining</nc>, by <nc>a computer</nc>, <nc>at least one frequency</nc> within <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc> <nc>that</nc> has <nc>the highest maximum intensity value</nc> among <nc>the at least a portion</nc> of <nc>the plurality</nc> of <nc>frequencies</nc>, and <nc>(c) determining</nc>, by <nc>a computer</nc>, <nc>a level</nc> of <nc>correlation</nc> between <nc>changes</nc> in <nc>intensity values</nc> during <nc>the time periods</nc> of <nc>a first frequency</nc> and <nc>changes</nc> in <nc>intensity values</nc> during <nc>the time periods</nc> of <nc>a second frequency</nc>; and outputting, by <nc>a computer</nc>, <nc>an indicator</nc> of <nc>the personality profile</nc> of <nc>the speaker</nc> based upon <nc>a combination</nc> of <nc>the calculated level</nc> of <nc>the survival element</nc> of <nc>the speaker</nc>, <nc>the calculated level</nc> of <nc>the homeostasis element</nc> of <nc>the speaker</nc>, and <nc>the calculated level</nc> of <nc>the growth element</nc> of <nc>the speaker</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the calculated level</nc> of <nc>the survival element</nc> comprises one of <nc>three possible levels</nc>, wherein <nc>the calculated level</nc> of <nc>the homeostasis element</nc> comprises one of <nc>three possible levels</nc>, wherein <nc>the calculated level</nc> of <nc>the growth element</nc> comprises one of <nc>three possible levels</nc>, and wherein <nc>the indicator</nc> of <nc>the personality profile</nc> comprises one of <nc>twenty-seven possible combinations</nc> of <nc>the levels</nc> of <nc>the survival element</nc>, <nc>the homeostasis element</nc> and <nc>the growth element</nc>.
|
7716576
|
10152453
|
1. <nc>A method</nc> of processing Extensible Markup Language <nc>(XML</nc>) documents, said <nc>method</nc> comprising: parsing <nc>an XML document</nc> comprising <nc>content</nc> in <nc>an XML format</nc>; converting said content into <nc>pcodes</nc> according to <nc>a conversion key</nc>, wherein said <nc>conversion key</nc> comprises <nc>a lookup table</nc> (<nc>LUT</nc>) comprising <nc>a plurality</nc> of <nc>XML tags</nc>, said <nc>XML tags</nc> <nc>each</nc> having <nc>a corresponding pcode</nc>, wherein <nc>an XML tag</nc> is converted into <nc>a pcode</nc> according to said <nc>conversion key</nc> and wherein said <nc>content</nc> converted into <nc>pcodes</nc> is convertible back to <nc>XML</nc> using said <nc>conversion key</nc>; representing <nc>a recurring sequence</nc> of <nc>different XML source code segments</nc> comprising <nc>multiple XML tags</nc> as <nc>a single same pcode</nc>, wherein said <nc>LUT</nc> further comprises <nc>an entry</nc> comprising said <nc>recurring sequence</nc> and said <nc>single pcode</nc>; generating <nc>a pcode file</nc> comprising <nc>said XML document</nc> parsed and converted into <nc>pcode</nc>, wherein said <nc>pcode file comprises</nc> said <nc>single pcode</nc> in <nc>lieu</nc> of <nc>each occurrence</nc> of <nc>said recurring sequence</nc> of <nc>XML tags</nc>; and forwarding said conversion key with <nc>said pcode file</nc> from <nc>one computer system</nc> to <nc>another computer system</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , wherein said <nc>single same pcode</nc> comprises <nc>three or more XML tags</nc>.
|
8533673
|
12052601
|
1. <nc>A method</nc> for creating <nc>a software program</nc> <nc>that</nc> integrates <nc>multiple programming languages</nc>, <nc>the method</nc> comprising: adding <nc>computer code</nc> written in <nc>a host language</nc> to <nc>a source code file</nc>, <nc>the source code file</nc> being stored on <nc>a computer-readable storage medium</nc>; adding <nc>inset computer code</nc> to <nc>the source code file</nc>, <nc>the inset computer code</nc> being written in <nc>a domain specific language</nc> <nc>that</nc> is different from <nc>the host language</nc>; and providing <nc>a domain specific language specification</nc> <nc>that</nc> comprises <nc>instructions</nc> written in <nc>the host language</nc> for executing <nc>the inset computer code</nc> in <nc>accordance</nc> with <nc>the domain specific language</nc>, <nc>the inset computer code</nc> being executed by relating <nc>tokens</nc> <nc>that</nc> include <nc>strings</nc> of <nc>characters</nc> from <nc>the domain specific language</nc> to <nc>corresponding tokens</nc> <nc>that</nc> include <nc>strings</nc> of <nc>characters</nc> from <nc>the host language</nc> and relating <nc>at least one grammatical rule</nc> for <nc>operations</nc> on <nc>tokens</nc> from <nc>the domain specific language</nc> to <nc>at least one corresponding grammatical rule</nc> for <nc>operations</nc> on <nc>tokens</nc> from <nc>the host language</nc>.
| 10
|
10. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the instructions</nc> operate to translate <nc>the inset computer code</nc> from <nc>a domain</nc> corresponding to <nc>the domain specific language</nc> to <nc>a domain</nc> corresponding to <nc>the host language</nc>.
|
10148660
|
15621986
|
1. <nc>A method</nc> for delivering <nc>author specific content</nc>, comprising: identifying <nc>any content</nc> on <nc>multiple online resources</nc> <nc>that</nc> appears to have been generated by <nc>a specific identified author</nc>; determining whether to authenticate <nc>authorship</nc> of <nc>the content</nc> and selectively authenticating <nc>content</nc> <nc>that</nc> appears to have been generated by <nc>the specific identified author</nc> as actually being generated by <nc>that author</nc>; delivering to <nc>a subscribing user</nc>, via <nc>an activity stream</nc>, <nc>an indication</nc> of <nc>the content</nc> identified as being by <nc>the specific identified author</nc> from <nc>the multiple online resources</nc>; and storing <nc>an author database</nc> <nc>that</nc> stores <nc>an identification</nc> of <nc>a first author</nc> along with <nc>a string</nc> provided by <nc>a content site</nc> having <nc>content</nc> generated by <nc>the first author</nc>, <nc>the string</nc> identifying <nc>a key</nc> <nc>that</nc> encrypts <nc>the first author's identification</nc> and <nc>the Universal Resource Locator</nc> (<nc>URL</nc>) of <nc>the content site</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein authenticating <nc>the content</nc> includes using <nc>a digital signature</nc> to authenticate <nc>the content</nc> as being generated by <nc>the specific identified author</nc>.
|
9251217
|
13755919
|
1. A method, comprising: receiving <nc>a search query</nc> from <nc>a member</nc> of <nc>a social network</nc>; identifying <nc>two or more categories</nc> of <nc>data</nc> within <nc>the social network</nc> <nc>that</nc> include <nc>information</nc> satisfying <nc>the received search query</nc>; ranking <nc>the identified two or more categories</nc> of <nc>data</nc> based on <nc>a selection criteria</nc> associated with <nc>the member</nc> of <nc>the social network</nc> and including <nc>a quality metric</nc> assigned to <nc>the information</nc> satisfying <nc>the received search query</nc> within <nc>the categories</nc> of <nc>data</nc>, the quality metric including: <nc>a percentage</nc> of <nc>users</nc> <nc>who</nc> have previously selected <nc>the at least one search query suggestion</nc>; and <nc>a percentage</nc> of <nc>users</nc> <nc>who</nc> have previously selected displayed <nc>results</nc> associated with <nc>the search query</nc>; and presenting <nc>at least one search query suggestion</nc> based on <nc>the ranking</nc> of <nc>the identified categories</nc> of <nc>data</nc> within <nc>the social network</nc>.
| 11
|
11. <nc>The method</nc> of <nc>claim</nc> 1 , wherein identifying <nc>two or more categories</nc> of <nc>data</nc> <nc>that</nc> include <nc>information</nc> satisfying <nc>the received search query</nc> includes <nc>a jobs category</nc> of <nc>data</nc> and <nc>a posts</nc> <nc>category</nc> of <nc>data</nc>.
|
8645418
|
13465465
|
1. <nc>A word mining</nc> and evaluating <nc>method</nc>, <nc>the method</nc> comprising: calculating <nc>a Document Frequency</nc> (<nc>DF</nc>) of <nc>a word</nc> in <nc>mass categorized data</nc>; evaluating <nc>the word</nc> in <nc>multiple single-aspects</nc> according to <nc>the DF</nc> of <nc>the word</nc>; and evaluating <nc>the word</nc> in <nc>a multiple-aspect</nc> according to <nc>the evaluations</nc> in <nc>the multiple single-aspects</nc> to obtain <nc>an importance weight</nc> of <nc>the word</nc>; wherein <nc>the</nc> evaluating <nc>the word</nc> in <nc>a multiple-aspect</nc> according to <nc>the evaluations</nc> in <nc>the multiple single-aspects</nc> to obtain <nc>the importance weight</nc> of <nc>the word</nc> comprises, classifying <nc>candidate words</nc> into <nc>levels</nc> according to <nc>DFs</nc> of <nc>the candidate words</nc>, wherein <nc>the levels</nc> comprises <nc>a SuperHigh level</nc>, <nc>a MidHigh level</nc>, <nc>a MidLow level</nc> and <nc>a SuperLow level</nc>; and for <nc>each candidate word</nc> in <nc>the SuperHigh level</nc>, the MidHigh level or the MidLow level, determining <nc>the importance weight</nc> of <nc>the candidate word</nc> according to, <nc>an absolute value</nc> of <nc>a difference</nc> between <nc>an average inverse document frequency</nc> (<nc>AVAIDF</nc>) and <nc>an inverse document frequency</nc> (<nc>IDF</nc>) of <nc>the candidate word</nc>, <nc>a linear combination</nc> of <nc>mutual information</nc> (<nc>MI</nc>), expect <nc>cross entropy</nc> (<nc>ECE</nc>) and <nc>entropy</nc> (<nc>ENT</nc>) of <nc>the candidate word</nc>, <nc>a combination</nc> of <nc>logarithmic normalized chi-square and information gain</nc> (<nc>IG</nc>) of <nc>the candidate word</nc>, and logarithmic normalized <nc>selective preference</nc> (<nc>SELPRE</nc>) of <nc>the candidate word</nc>; and for <nc>each candidate word</nc> in <nc>the SuperLow level</nc>, determining <nc>the importance weight</nc> of <nc>the candidate word</nc> according to, <nc>an absolute value</nc> of <nc>a difference</nc> between <nc>an average inverse document frequency</nc> (<nc>AVAIDF</nc>) and <nc>an inverse document frequency</nc> (<nc>IDF</nc>) of <nc>the candidate word</nc>, <nc>a linear combination</nc> of <nc>mutual information</nc> (<nc>MI</nc>), expect <nc>cross entropy</nc> (<nc>ECE</nc>) and <nc>entropy</nc> (<nc>ENT</nc>) of <nc>the candidate word</nc>, and <nc>a combination</nc> of <nc>logarithmic normalized chi-square and information gain</nc> (<nc>IG</nc>) of <nc>the candidate word</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the importance weight</nc> of <nc>the candidate word</nc> in <nc>the SuperHigh level</nc> or <nc>the MidHigh level</nc> is determined according to <nc>a following formula</nc>:
<nc>SuperHigh( w</nc> )=<nc>MidHigh( w</nc> )=<nc>Diff</nc>( <nc>w</nc> )*<nc>ProbBased</nc>( <nc>w</nc> )*<nc>Prob DF Rel</nc>( <nc>w</nc> )*<nc>SelPre</nc>( <nc>w</nc> ); <nc>the importance weight</nc> of <nc>the candidate word</nc> in <nc>the MidLow level</nc> is determined according to <nc>a following formula</nc>:
<nc>MidLow</nc>( <nc>w</nc> )=<nc>Diff</nc>( <nc>w</nc> )*ProbBased( <nc>w</nc> )*<nc>Prob DF Rel</nc>( <nc>w</nc> )+<nc>SelPre</nc>( <nc>w</nc> ); wherein, SuperHigh(w) denotes <nc>the importance weight</nc> of <nc>the candidate word</nc> in <nc>the SuperHigh level</nc>, <nc>MidHigh(w</nc>) denotes <nc>the importance weight</nc> of <nc>the candidate word</nc> in <nc>the MidHigh level</nc>, <nc>MidLow(w</nc>) denotes <nc>the importance weight</nc> of <nc>the candidate word</nc> in <nc>the MidLow level</nc>, <nc>Diff</nc> (<nc>w</nc>) denotes <nc>the absolute value</nc> of <nc>the difference</nc> between <nc>the average inverse document frequency</nc> and <nc>the inverse document frequency</nc> of <nc>the candidate word</nc>, <nc>ProbBased(w</nc>) denotes <nc>the linear combination</nc> of <nc>the mutual information</nc>, <nc>the</nc> expect <nc>cross entropy</nc> and <nc>the entropy</nc> of <nc>the candidate word</nc>, <nc>ProbDFRel(w</nc>) denotes <nc>the combination</nc> of <nc>the logarithmic normalized chi-square</nc> and <nc>the information gain</nc> of <nc>the candidate word</nc>, and <nc>SelPre(w</nc>) denotes <nc>the logarithmic</nc> normalized selective preference of <nc>the candidate word</nc>.
|
9443139
|
14587858
|
1. <nc>A method</nc> of obtaining <nc>information</nc> of <nc>interest</nc> from <nc>a binarized document</nc>, <nc>the method</nc> comprising: storing <nc>information</nc> associating <nc>a plurality</nc> of <nc>different label aliases</nc> with <nc>a first label</nc> for <nc>first information</nc> of <nc>interest</nc>; storing in <nc>a nodal structure</nc>, <nc>information</nc> representing <nc>an expected relationship</nc> between <nc>objects</nc> of <nc>a first label alias</nc> in said <nc>plurality</nc> of <nc>different label aliases</nc>, <nc>individual objects</nc> of said <nc>first label alias</nc> being linked in said <nc>nodal structure</nc> to <nc>other objects</nc>, <nc>each</nc> of <nc>the linked objects</nc> being <nc>a character</nc> or <nc>character string</nc>, <nc>individual links</nc> in said <nc>nodal structure</nc> having <nc>a probability</nc> associated with <nc>the individual link</nc>; scoring <nc>a first portion</nc> of <nc>the binarized document</nc> against said <nc>nodal structure</nc> to determine <nc>scores</nc> for <nc>multiple label aliases</nc>; determining based on said <nc>generated scores</nc> if said <nc>first label alias</nc> is present in <nc>said first portion</nc> of <nc>the binarized document</nc>; and in <nc>response</nc> to determining <nc>that</nc> said <nc>first label alias</nc> is present in <nc>said first portion</nc> of <nc>the binarized document</nc>, extracting <nc>information</nc> from <nc>the first portion document</nc> corresponding to said <nc>first label alias</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining based on said generated <nc>scores</nc> if <nc>a second label alias</nc> is present in <nc>said first portion</nc> of <nc>the binarized document</nc>; and in <nc>response</nc> to determining <nc>that</nc> said <nc>second label alias</nc> is present in <nc>said first portion</nc> of <nc>the binarized document</nc>, extracting <nc>information</nc> from <nc>the first portion</nc> of <nc>the binarized document</nc> corresponding to said <nc>second label alias</nc>.
|
9633173
|
14202331
|
1. <nc>A method</nc> implemented on <nc>a processor</nc>, <nc>the method</nc> comprising: receiving, on <nc>a computer host</nc>, <nc>an electronic document</nc> representing <nc>a pharmaceutical prescription</nc>; identifying <nc>constituent regions</nc> <nc>that</nc> include <nc>at least a first portion</nc> and <nc>a second portion</nc> within <nc>the electronic document</nc>; identifying <nc>first spatial frequencies</nc> for <nc>the first portion</nc> within <nc>the electronic document</nc>; identifying <nc>second spatial frequencies</nc> for <nc>the second portion</nc> within <nc>the electronic document</nc>; identifying <nc>a header</nc> based upon <nc>the first spatial frequencies</nc>, wherein identifying <nc>the first and second spatial frequencies</nc> includes performing <nc>a fast Fourier</nc> Transform on <nc>a portion</nc> of <nc>a document</nc> and translating <nc>spatial information</nc> into <nc>frequency information</nc> such that (i) <nc>the header</nc> identifying <nc>the prescriber</nc> is identified from <nc>the first spatial frequencies</nc>; and (ii) <nc>the second spatial frequencies</nc> are analyzed using <nc>the profile</nc> of <nc>the identified prescriber</nc>; using <nc>the header</nc> to identify <nc>a prescriber</nc>; retrieving <nc>a profile</nc> specific for <nc>the prescriber</nc>; analyzing <nc>the second spatial frequencies</nc> using <nc>the profile</nc> specific for <nc>the prescriber</nc> such that <nc>the second spatial frequencies</nc> are compared to <nc>spatial frequency domain information</nc> from <nc>the profile</nc> specific for <nc>the prescriber</nc>, the prescriber-specific profile constructed from <nc>the prescriber's past records</nc> and including <nc>isolated handwriting</nc> of <nc>the prescriber</nc> in <nc>a spatial frequency domain</nc>; and creating, based upon <nc>results</nc> from analyzing <nc>the second spatial frequencies</nc> using <nc>the profile</nc> for <nc>the prescriber</nc>, <nc>a transaction record</nc> on <nc>the computer host</nc> for <nc>a medical transaction</nc> associated with <nc>the pharmaceutical prescription</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , wherein retrieving <nc>the profile</nc> for <nc>the prescriber</nc> includes retrieving <nc>a dictionary</nc> of <nc>terms</nc> for <nc>the prescriber</nc> represented in <nc>the spatial frequency domain</nc>.
|
7752195
|
11506495
|
1. <nc>A computer-implemented method</nc> for generating <nc>a universal query</nc> result <nc>set</nc>, comprising: under <nc>control</nc> of <nc>one or more computer systems</nc> configured with <nc>executable instructions</nc>, submitting <nc>a received query</nc> to <nc>a plurality</nc> of <nc>search indexes</nc>, <nc>each search index</nc> associated with <nc>a relevance ranking function</nc> for <nc>ranking items</nc> according to <nc>an individual ranking scale</nc>, at least two of <nc>the individual ranking scales</nc> being different; receiving from <nc>the plurality</nc> of <nc>search indexes</nc> <nc>a plurality</nc> of <nc>search index result sets</nc>, <nc>each search index</nc> result set including <nc>one or more items</nc> related to <nc>the received query</nc> and ranked according to one of <nc>the individual ranking scales</nc>; determining <nc>a query index association</nc> for <nc>each</nc> of <nc>the plurality</nc> of <nc>search indexes</nc>; and including in <nc>the universal query</nc> result set <nc>a list</nc> of <nc>items</nc> obtained from <nc>different ones</nc> of <nc>the plurality</nc> of <nc>search index result sets</nc>, <nc>the list</nc> being ordered based at least in <nc>part</nc> on <nc>a probability</nc> of <nc>the items</nc> satisfying <nc>the query</nc>, <nc>the probability</nc> of <nc>the items</nc> satisfying <nc>the query</nc> determined at least in <nc>part</nc> by: computing <nc>an allocation score</nc> for <nc>each</nc> of <nc>the plurality</nc> of <nc>search indexes</nc>, the allocation score for <nc>each search index</nc> being computed based at least in <nc>part</nc> upon <nc>a relative number</nc> of <nc>search results</nc> in <nc>the corresponding search index result set</nc>; for <nc>each</nc> of <nc>the plurality</nc> of <nc>received search index result sets</nc>, computing <nc>a universal item score</nc> for <nc>at least one item</nc> identified in <nc>the search index result set</nc>, the universal item score for <nc>each identified item</nc> being computed at least in <nc>part</nc> by normalizing <nc>the individual ranking scale</nc> for <nc>each search index result</nc> set to <nc>a common ranking scale</nc>; and determining <nc>a probability</nc> of <nc>each item</nc> satisfying <nc>the query</nc> based at least on <nc>a combination</nc> of <nc>the allocation score</nc> for <nc>the search index</nc> associated with <nc>the item</nc> and <nc>the universal item score</nc> for <nc>the item</nc>.
| 12
|
12. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , wherein <nc>the allocation score</nc> and <nc>the universal item score</nc> are combined by at least multiplying <nc>the allocation score</nc> and <nc>the universal item score</nc>.
|
8122042
|
12493105
|
1. <nc>A method</nc> of determining <nc>relevance</nc> of <nc>a content identifier</nc>, comprising: receiving, at <nc>a direct answer computer system</nc>, <nc>a search query</nc> over <nc>a network</nc>; determining, at <nc>the direct answer computer system</nc>, <nc>one or more answer entities</nc> from <nc>one or more answer candidate snippets</nc>, wherein <nc>an answer candidate snippet</nc> comprises <nc>at least a portion</nc> of <nc>content</nc> available over <nc>the network</nc> for <nc>an answer candidate</nc>; determining, at <nc>the direct answer computer system</nc>, <nc>a content identifier</nc> for <nc>an answer candidate</nc>; determining <nc>a popularity</nc> for <nc>the content identifier</nc> and adjusting <nc>an indicator</nc> of <nc>the relevance</nc> for <nc>the content identifier</nc> in <nc>accordance</nc> with <nc>the popularity</nc> of <nc>the content identifier</nc>, wherein <nc>the popularity</nc> of <nc>the content identifier</nc> comprises <nc>a click count</nc> for <nc>the content identifier</nc>, <nc>a click count</nc> for <nc>a main web site</nc> for <nc>the content identifier</nc>, <nc>a count</nc> of <nc>references</nc> to <nc>the content identifier</nc>, and a count of <nc>references</nc> to <nc>the main web site</nc>; tokenizing, at <nc>the direct answer computer system</nc>, <nc>a title</nc> for <nc>the content identifier</nc>; and performing <nc>a comparison</nc>, at <nc>the direct answer computer system</nc>, between <nc>a vector</nc> of <nc>tokens</nc> for <nc>the title</nc> and <nc>a vector</nc> of <nc>the one or more answer entities</nc>; adjusting <nc>an indicator</nc> of <nc>the relevance</nc> for <nc>the content identifier</nc> in <nc>accordance</nc> with <nc>the comparison</nc>; and sending <nc>at least one answer candidate snippet</nc> for <nc>a response</nc> to <nc>the search query</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>an indicator</nc> of <nc>the relevance</nc> for <nc>the content identifier</nc> is <nc>a score</nc> for <nc>the content identifier</nc> <nc>that</nc> is increased when <nc>similar tokens</nc> from <nc>the title</nc> are found in <nc>the answer entities</nc> from <nc>the one or more answer candidate snippets</nc>.
|
9015043
|
12896507
|
1. <nc>A computer-implemented method</nc> comprising: receiving <nc>an electronic representation</nc> of <nc>a conversation</nc> between <nc>a plurality</nc> of <nc>human voices</nc>; recognizing <nc>one or more words</nc> in <nc>a first portion</nc> of <nc>the electronic representation</nc> of <nc>the conversation</nc>; recognizing <nc>one or more words</nc> in <nc>a second portion</nc> of <nc>the electronic representation</nc> of <nc>the conversation</nc>, wherein <nc>the second portion</nc> is different from <nc>the first portion</nc>, wherein <nc>the first portion</nc> has <nc>a duration</nc> from <nc>a first time</nc> to <nc>a second time</nc>, and wherein <nc>the second portion</nc> has <nc>a duration</nc> from <nc>a third time</nc> to <nc>a fourth time</nc>, <nc>the third time</nc> being after <nc>the first time</nc> and before <nc>the second time</nc>; selecting <nc>search terms</nc> for <nc>a search query</nc>, wherein <nc>the search terms</nc> include: <nc>at least one word</nc> from <nc>the one or more recognized words</nc> in <nc>the first portion</nc> <nc>that</nc> is selected as <nc>a search term</nc> based on whether <nc>the word</nc> is <nc>a proper noun</nc>, based on <nc>an inverse</nc> of <nc>a frequency</nc> of <nc>the word</nc> in <nc>a corpus</nc> of <nc>documents</nc>, and based on <nc>a number</nc> of <nc>times</nc> that <nc>the word</nc> is recognized in <nc>the first portion</nc> of <nc>the electronic representation</nc> of <nc>the conversation</nc>, <nc>at least one word</nc> <nc>that</nc> was not spoken in <nc>the conversation</nc> and is related to <nc>a word</nc> <nc>that</nc> was spoken in <nc>the conversation</nc>, and <nc>one or more words</nc> recognized in <nc>the second portion</nc> of <nc>the electronic representation</nc> of <nc>the conversation</nc>; causing <nc>the search terms</nc> to be displayed on <nc>a display device</nc> in <nc>a text format</nc>; receiving <nc>a search query</nc> <nc>that</nc> includes at least one of <nc>the search terms</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the search terms</nc> are displayed in <nc>a format</nc> <nc>that</nc> includes <nc>indicia</nc> used to convey <nc>a significance</nc> of <nc>each search term</nc>.
|
9495458
|
14567324
|
1. <nc>A method</nc> comprising: identifying, at <nc>a server</nc>, based on <nc>keywords</nc> identified by <nc>an intercept module monitoring user input and user communications</nc>, <nc>items</nc> of <nc>interest</nc> to present to <nc>a user</nc> via <nc>a webpage</nc>, <nc>each</nc> of <nc>the keywords</nc> having <nc>a date</nc> and <nc>time stamp</nc> and ranked according to <nc>occurrence</nc> in <nc>the user input</nc> and <nc>the user communications</nc>; reducing <nc>a ranking</nc> of one of <nc>the keywords</nc> identified by <nc>the intercept module</nc> based on <nc>expiration</nc> of <nc>a user defined period</nc> of <nc>time</nc> <nc>that</nc> begins on <nc>a date</nc> and a time identified by <nc>a time stamp</nc> of the one of <nc>the keywords</nc>; and generating <nc>the webpage</nc> including <nc>information</nc> related to at least one of <nc>the items</nc> of <nc>interest</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: receiving <nc>the keywords</nc> from <nc>a user device</nc> associated with <nc>the user</nc>, wherein <nc>the keywords</nc> received are automatically identified and ranked at <nc>the user device</nc>.
|
9444914
|
14027257
|
1. <nc>A system</nc>, comprising: <nc>a processor</nc>; and <nc>packet controller logic circuits</nc> coupled to <nc>the processor</nc> and implementing: <nc>a distribution module</nc>; <nc>multiple interleaved sequences</nc> of <nc>configurable parsing engines</nc> and concatenating <nc>modules</nc> forming <nc>configurable parsing engine</nc> and concatenating <nc>module pairs</nc>; and <nc>an aggregation module</nc>; wherein <nc>the distribution module</nc> is arranged to distribute <nc>an information unit</nc> between <nc>the multiple interleaved sequences</nc> of <nc>configurable parsing engines</nc> and concatenating <nc>modules</nc>, and apply <nc>a load balancing scheme</nc> in <nc>order</nc> to distribute <nc>the information unit</nc>, wherein <nc>different portions</nc> of <nc>the information unit</nc> are processed by <nc>different configurable parsing engines</nc>, wherein <nc>at least one configurable parsing engine</nc> is arranged to process <nc>a portion</nc> of <nc>the information unit</nc> in <nc>response</nc> to <nc>at least a previous processing result</nc> provided from <nc>a previous configurable parsing engine</nc>, and to generate <nc>a current processing result</nc> to be used by <nc>a next configurable parsing engine</nc>, wherein <nc>the multiple interleaved sequences</nc> include <nc>a first interleaved sequence</nc> having <nc>at least two configurable parsing engine</nc> and concatenating <nc>module pairs</nc>, and wherein <nc>at least a portion</nc> of <nc>the information unit</nc> is processed sequentially by <nc>the at least two configurable parsing engine</nc> and concatenating <nc>module pairs</nc>; and wherein <nc>the aggregation module</nc> collects <nc>outputs</nc> from <nc>the multi-interleaved sequences</nc>.
| 6
|
6. <nc>The system</nc> according to <nc>claim</nc> 1 , wherein <nc>one or more configurable parsing engines</nc> is arranged to generate <nc>a current processing result</nc> to be used by <nc>another module</nc> of <nc>the system</nc>.
|
8589150
|
12985272
|
1. <nc>A computer program product</nc>, comprising <nc>a non-transitory computer usable medium</nc> having <nc>a computer readable program code</nc> embodied therein, <nc>the computer readable program code</nc> adapted to be executed to implement <nc>a method</nc> for dynamically correcting <nc>grammar</nc> associated with <nc>text</nc>, <nc>the method</nc> comprising: receiving <nc>a request</nc> from <nc>a user</nc> to change <nc>text</nc> within <nc>an application</nc>; changing <nc>all instances</nc> of <nc>the text</nc> within <nc>the application</nc>, according to <nc>the request</nc>; and dynamically correcting <nc>grammar</nc> associated with <nc>the changed text</nc> when <nc>the changed text</nc> is displayed to <nc>the user</nc>, including dynamically correcting <nc>the display</nc> of <nc>additional text</nc> associated with <nc>the changed text</nc>, according to <nc>one or more rules</nc> associated with <nc>a language</nc> of <nc>the changed text</nc>.
| 7
|
7. <nc>The computer program product</nc> of <nc>claim</nc> 1 , wherein <nc>the request</nc> to change <nc>the text</nc> includes <nc>a request</nc> to change <nc>the wording</nc> of <nc>the text</nc> from <nc>a word</nc> starting with <nc>a vowel</nc> to <nc>a word</nc> starting with <nc>a consonant</nc>.
|
8543373
|
12270082
|
1. <nc>A computer system</nc> for managing <nc>word usage frequencies</nc>, <nc>the computer system</nc> comprising: <nc>one or more processors</nc>, <nc>one or more computer-readable memories</nc> and <nc>one or more computer-readable tangible storage devices</nc>; <nc>program instructions</nc>, stored on at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by at least one of <nc>the one or more processors</nc> via at least one of <nc>the one or more memories</nc>, to receive: <nc>an identifier</nc> of <nc>a location</nc>; <nc>an identifier</nc> of <nc>a number</nc> of <nc>document levels</nc> at <nc>the location</nc>, <nc>the document levels</nc> comprising <nc>at least one document</nc>; and <nc>an identifier</nc> of <nc>a minimum number</nc> of <nc>words</nc>; <nc>program instructions</nc>, stored on the at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by the at least one of <nc>the one or more processors</nc> via the at least one of <nc>the one or more memories</nc>, to determine whether <nc>the minimum number</nc> of <nc>words</nc> are present in <nc>the at least one document</nc>; <nc>program instructions</nc>, stored on the at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by the at least one of <nc>the one or more processors</nc> via the at least one of <nc>the one or more memories</nc>, responsive to determining that <nc>the minimum number</nc> of <nc>words</nc> are present in <nc>the at least one document</nc>, to analyze <nc>all</nc> of <nc>the at least one document</nc> to determine whether <nc>the at least one document</nc> comprises <nc>Latin based words</nc> or <nc>Sino-Tibetan based words</nc>; <nc>program instructions</nc>, stored on the at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by the at least one of <nc>the one or more processors</nc> via the at least one of <nc>the one or more memories</nc>, responsive to determining that <nc>the at least one document</nc> comprises <nc>the Latin based words</nc>, to populate <nc>a Latin based word list</nc> with <nc>unique Latin based words</nc> in <nc>the at least one document</nc> and to determine <nc>a frequency</nc> of <nc>each</nc> of <nc>the unique Latin based words</nc> in <nc>the at least one document</nc>; <nc>program instructions</nc>, stored on the at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by the at least one of <nc>the one or more processors</nc> via the at least one of <nc>the one or more memories</nc>, responsive to determining that <nc>the at least one document</nc> comprises <nc>the Sino-Tibetan based words</nc>, to populate <nc>a Sino-Tibetan based word list</nc> with <nc>unique Sino-Tibetan based words</nc> in <nc>the at least one document</nc> and to determine <nc>a frequency</nc> of <nc>the unique Sino-Tibetan based words</nc> in <nc>the at least one document</nc>; and <nc>program instructions</nc>, stored on the at least one of <nc>the one or more storage devices</nc> for <nc>execution</nc> by the at least one of <nc>the one or more processors</nc> via the at least one of <nc>the one or more memories</nc>, to generate <nc>results</nc> comprising <nc>the unique Latin based words</nc>, <nc>the frequency</nc> of <nc>the unique Latin based words</nc>, <nc>the unique Sino-Tibetan based words</nc>, and <nc>the frequency</nc> of <nc>the unique Sino-Tibetan based words</nc>.
| 3
|
3. <nc>The computer system</nc> of <nc>claim</nc> 1 , wherein <nc>the location</nc> is <nc>a website</nc> having <nc>hyperlinks</nc>.
|
7895531
|
11151686
|
1. A method of providing <nc>a floating command object</nc> <nc>that</nc> is contextually relevant to <nc>selected text</nc>, <nc>the method</nc> comprising: upon receiving <nc>a selection</nc> of <nc>text</nc> in <nc>an electronic document</nc> for <nc>editing</nc>, displaying <nc>a command object</nc> adjacent to <nc>the selected text</nc> such that <nc>at least a portion</nc> of <nc>the selected text</nc> remains visible, <nc>the command</nc> object providing <nc>text editing functionality</nc> in <nc>response</nc> to <nc>the selection</nc> of <nc>the text</nc>; displaying in <nc>the command</nc> object <nc>a set</nc> of <nc>functionality commands</nc> <nc>that</nc> are relevant to editing <nc>the selected text</nc>, <nc>the set</nc> of <nc>functionality commands</nc> being <nc>a subset</nc> of <nc>a broader range</nc> of <nc>functionality</nc> commands available for editing <nc>the selected text</nc>, wherein displaying in <nc>the command</nc> object <nc>the set</nc> of <nc>functionality</nc> commands comprises displaying <nc>the set</nc> of <nc>functionality</nc> commands with <nc>a first set</nc> of <nc>visual representations</nc> similar to <nc>a second set</nc> of <nc>visual representations</nc> associated with displaying <nc>the broader range</nc> of <nc>functionality commands</nc>; associating <nc>an opacity</nc> of <nc>the displayed command object</nc> to <nc>a distance</nc> between <nc>an electronic pointer</nc> and <nc>the displayed command object</nc>; and continuing to display <nc>the command object</nc> after receiving <nc>a selection</nc> of one of <nc>the set</nc> of <nc>functionality commands</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein displaying in <nc>the command</nc> object <nc>the set</nc> of <nc>functionality commands</nc> <nc>that</nc> are relevant to editing <nc>the selected text</nc> includes displaying <nc>the set</nc> of <nc>functionality</nc> <nc>commands</nc> for applying <nc>a formatting property</nc> to <nc>the selected text</nc>.
|
8271600
|
13269133
|
1. <nc>A computer-implemented method</nc> for facilitating <nc>online communication</nc> among <nc>multiple online messaging identities</nc>, <nc>the method</nc> comprising: identifying <nc>a target online messaging identity</nc> <nc>who</nc> is accessible over <nc>a network</nc> of <nc>computers</nc>, and <nc>who</nc> desires <nc>interaction</nc> with <nc>other online messaging identities</nc> <nc>who</nc> are presently subscribed to <nc>an online messaging service</nc>; presenting <nc>questions</nc> to <nc>the target online messaging identity</nc>, at least one of <nc>the questions</nc> including <nc>selectable responses</nc> associated with <nc>the question</nc>; receiving, from <nc>the target online messaging identity</nc>, responses to <nc>the questions</nc>; accessing, from <nc>computer-accessible memory</nc>, stored <nc>responses</nc> to <nc>the questions</nc>, at least one of <nc>the stored responses</nc> having been received from <nc>another online messaging identity</nc> presently subscribed to <nc>the online messaging service</nc>; comparing <nc>responses</nc> received from <nc>the target online messaging identity</nc> to <nc>the accessed stored responses</nc>; identifying, based at least in <nc>part</nc> on <nc>results</nc> of <nc>the comparison</nc>, a group of less than <nc>all</nc> of <nc>online messaging identities</nc> <nc>who</nc> can be contacted through <nc>the online messaging service</nc>, as being of <nc>interest</nc> to <nc>the target online messaging identity</nc>; and introducing <nc>the identified group</nc> of <nc>online messaging identities</nc> <nc>who</nc> can be contacted through <nc>the online messaging service</nc> to <nc>the target online messaging identity</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 wherein: comparing <nc>responses</nc> received to <nc>accessed stored responses</nc> comprises converting <nc>a received response</nc> into <nc>a representative code</nc> and comparing <nc>the representative code</nc> to stored <nc>codes</nc> corresponding to and representative of <nc>responses</nc> having been received from <nc>the other instant messaging identity</nc>.
|
7728826
|
10597201
|
1. <nc>A mobile display apparatus</nc> serving as <nc>a client device</nc> of <nc>an external host apparatus</nc>, <nc>the mobile display apparatus</nc> comprising: <nc>a display section</nc> including <nc>a plurality</nc> of scanning <nc>lines</nc> arranged in <nc>a row</nc> and <nc>a plurality</nc> of <nc>signal lines</nc> arranged in <nc>a column</nc> and respectively intersecting with <nc>the scanning lines</nc>; <nc>a voice output section</nc>; <nc>a text code input section</nc> arranged to receive <nc>an input text code</nc> from <nc>the external host apparatus</nc>; <nc>a video signal input section</nc> arranged to receive <nc>an input video signal</nc> from <nc>the external host apparatus</nc>; <nc>a display control section</nc> arranged to display in <nc>the display section text</nc> corresponding to <nc>the input text code</nc> and including: <nc>a scanning line drive circuit</nc> defining <nc>a display section drive circuit</nc> <nc>that</nc> is arranged to drive <nc>the display section</nc> by sequentially supplying <nc>scanning signals</nc> to <nc>the scanning lines</nc>; and <nc>a signal line drive circuit</nc> arranged to supply <nc>video signals</nc> to <nc>the signal lines</nc>, <nc>the signal line drive circuit</nc> including <nc>a first signal line drive circuit</nc> <nc>that</nc> is arranged to receive <nc>the video signal</nc> from <nc>the video signal input section</nc> and <nc>a second signal line drive circuit</nc> <nc>that</nc> is arranged to receive <nc>a video signal</nc> corresponding to <nc>the text</nc> corresponding to <nc>the input text code</nc>; and <nc>a voice output control section</nc> arranged <nc>to output voice</nc> sounds corresponding to <nc>the input text code</nc> through <nc>the voice output section</nc>; wherein <nc>the text code input section</nc> outputs <nc>the input text code</nc> to <nc>the display control section</nc> and the voice output control section to display in <nc>the display section</nc> <nc>the text</nc> corresponding to <nc>the input text code</nc>, and to output <nc>the voice</nc> sounds corresponding to <nc>the input text code</nc> through <nc>the voice output section</nc>; <nc>the display control section displays</nc> in <nc>the display section</nc> <nc>both an image</nc> based on <nc>the input video signal</nc> supplied to <nc>the video signal input section</nc> and <nc>the text</nc> corresponding to <nc>the input text code</nc> so that <nc>the text</nc> is superimposed on <nc>the image</nc>; and <nc>the first signal line drive circuit</nc> and <nc>the second signal line drive circuit</nc> share <nc>the signal lines</nc>.
| 2
|
2. <nc>The mobile display apparatus</nc> as set forth in <nc>claim</nc> 1 , wherein: <nc>the display section</nc> includes <nc>a display element</nc> driven by <nc>a thin film element</nc>; and <nc>the text code input section</nc>, <nc>the display control section</nc>, and <nc>the voice output control section</nc> are either directly provided on <nc>a thin film substrate</nc> on <nc>which</nc> <nc>a pixel driving circuit element</nc> of <nc>the display element</nc> is provided, or include <nc>active elements</nc> provided on <nc>another substrate</nc> <nc>which</nc> is bonded to <nc>the thin film substrate</nc>.
|
9703766
|
14993988
|
1. <nc>A method</nc> for generating <nc>tables</nc> from <nc>print-ready digital source documents</nc>, comprising: receiving, by <nc>a processor</nc> of <nc>a computing device</nc>, <nc>a print-ready digital source document</nc>, <nc>the digital source document</nc> comprising <nc>at least one rendered page</nc>; identifying, by <nc>the processor</nc>, <nc>one or more text fragments</nc> in <nc>the at least one rendered page</nc>, <nc>each</nc> of <nc>the text fragments</nc> comprising <nc>text</nc>, <nc>spatial coordinates</nc> indicating <nc>the positioning</nc> of <nc>the text fragment</nc> on <nc>the rendered page</nc>, and an index assigned based on <nc>the spatial coordinates</nc> on <nc>the rendered page</nc>; generating, by <nc>the processor</nc>, <nc>a wrapping region collection</nc> comprising <nc>one or more wrapping regions</nc>, wherein <nc>each</nc> of <nc>the wrapping regions</nc> comprises <nc>one or more fragment runs</nc>, and wherein <nc>each</nc> of <nc>the one or more fragment runs</nc> comprises <nc>a subset</nc> of <nc>the one or more text fragments</nc> <nc>that</nc> are adjacent to one another and within <nc>a predetermined horizontal separation threshold</nc> and <nc>a vertical separation threshold</nc>; calculating, by <nc>the processor</nc>, for <nc>each</nc> of <nc>the one or more wrapping regions</nc> of <nc>the wrapping region collection</nc>, a tabular score, a narrative score, and a label score, wherein <nc>each</nc> of <nc>the tabular score</nc>, <nc>narrative score</nc>, and <nc>label score</nc> is <nc>a measure</nc> of <nc>qualification</nc> of <nc>a wrapping region</nc> as <nc>a tabular block type</nc>, as <nc>a narrative block type</nc>, or as <nc>a label block type</nc>, respectively; assigning, by <nc>the processor</nc>, a block type to <nc>each</nc> of <nc>the one or more wrapping regions</nc> based on <nc>the corresponding calculated tabular score</nc>, narrative score and label score; generating <nc>a wrapping region group</nc> set comprising <nc>one or more wrapping region groups</nc>, wherein <nc>each</nc> of <nc>the one or more wrapping region groups</nc> comprises <nc>a subset</nc> of <nc>the one or more wrapping regions</nc> <nc>that</nc> are spatially related to one another; generating, by <nc>the processor</nc>, a block set comprising <nc>one or more blocks</nc>, wherein <nc>each</nc> of <nc>the one or more blocks</nc> comprises <nc>a subset</nc> of <nc>the one or more wrapping region groups</nc> <nc>that</nc> are spatially related to one another; and generating, by <nc>the processor</nc>, <nc>one or more tables</nc>, <nc>each</nc> of <nc>the one or more tables</nc> comprising <nc>the text fragments</nc> corresponding to one of <nc>the one or more blocks</nc>, wherein <nc>each</nc> of <nc>the one or more tables</nc> comprises <nc>the corresponding text fragments</nc>, <nc>each</nc> organized into <nc>corresponding fields</nc> of <nc>the one or more tables</nc>; <nc>wherein the tabular score</nc>, <nc>the narrative score</nc> and <nc>the label score</nc> of <nc>each</nc> of <nc>the one or more wrapping regions</nc> are calculated based on <nc>one or more attributes</nc> selected from <nc>the group</nc> consisting of (<nc>i</nc>) <nc>a normalization ratio</nc>, (<nc>ii</nc>) <nc>a density ratio</nc>, (<nc>iii</nc>) <nc>an alignment ratio</nc>, (<nc>iv</nc>) <nc>a capital or non-alphabetic ratio</nc>, (v) <nc>a text fragment quantity</nc>, and <nc>(vi</nc>) <nc>a bold count</nc>; wherein (i) <nc>the normalization ratio</nc> is <nc>the number</nc> of <nc>normalized text fragments</nc> divided by <nc>the total number</nc> of <nc>text fragments</nc> within <nc>a wrapping region</nc>, <nc>(ii</nc>) <nc>the density ratio</nc> is <nc>the percentage</nc> of <nc>a bounding box area</nc> of <nc>a wrapping region</nc> occupied by <nc>text fragment bounding boxes</nc>, (iii) <nc>the alignment ratio</nc> is <nc>the number</nc> of <nc>normalized text fragments</nc> <nc>that</nc> fit within <nc>at least one alignment group</nc> divided by <nc>the total number</nc> of <nc>normalized text fragments</nc> within <nc>a wrapping region</nc>, (iv) <nc>the capital or non-alphabetic ratio</nc> is <nc>the number</nc> of <nc>normalized text fragments</nc> <nc>that</nc> start with <nc>either a capital letter</nc> or <nc>a non-alphabetic character</nc> divided by <nc>the number</nc> of <nc>normalized text fragments</nc>, (v) <nc>the text fragment quantity</nc> is <nc>a number</nc> of <nc>text fragments</nc> among <nc>the subset</nc> of <nc>the one or more text fragments</nc> corresponding to <nc>each</nc> of <nc>the one or more wrapping regions</nc>, and (vi) <nc>the bold count</nc> is <nc>the number</nc> of <nc>text fragments</nc> with <nc>bolded text</nc>.
| 11
|
11. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the print-ready digital source document</nc> is <nc>a fixed-layout file</nc>.
|
10140279
|
14995156
|
1. <nc>A method</nc>, comprising: at <nc>one or more computing devices</nc> comprising <nc>one or more processors</nc> and <nc>storage media</nc> storing <nc>one or more computer programs</nc> executed by <nc>the one or more processors</nc> to perform <nc>the method</nc>, performing <nc>operations</nc> comprising: in <nc>response</nc> to detecting <nc>a semantic organization event</nc> in <nc>a first graphical user interface content-view</nc> of <nc>a spreadsheet</nc>, determining whether <nc>semantic organization data</nc> associated with <nc>a first set</nc> of <nc>data cells</nc> of <nc>the spreadsheet</nc> should be stored, wherein <nc>the semantic organization data</nc> describes how <nc>the spreadsheet</nc> is organized; wherein <nc>the first set</nc> of <nc>data cells</nc> includes <nc>a plurality</nc> of <nc>columns</nc>; wherein <nc>the plurality</nc> of <nc>columns</nc> of <nc>the first set</nc> of <nc>data cells</nc> includes <nc>a plurality</nc> of <nc>data cell values</nc>; wherein <nc>the spreadsheet</nc> includes <nc>a header</nc> <nc>that</nc> collectively groups <nc>the plurality</nc> of <nc>columns</nc> within <nc>the spreadsheet</nc>; wherein <nc>the header</nc> and <nc>the plurality</nc> of <nc>data cell values</nc> are content input into <nc>the spreadsheet</nc>; wherein <nc>the semantic organization data</nc> includes <nc>the header</nc>; in <nc>response</nc> to determining that <nc>semantic organization data</nc> should be stored, storing <nc>semantic organization data</nc> associated with <nc>the first set</nc> of <nc>data cells</nc>; generating <nc>a preview thumbnail image</nc> based on <nc>the semantic organization data</nc>; displaying <nc>a graphical user interface</nc> <nc>semantic-view</nc> of <nc>the spreadsheet</nc>, wherein <nc>the graphical user</nc> interface <nc>semantic-view</nc> comprises <nc>the preview thumbnail image</nc> corresponding to <nc>the first set</nc> of <nc>data cells</nc>; wherein <nc>the preview thumbnail image</nc> displayed includes <nc>display</nc> of <nc>the header</nc> <nc>that</nc> collectively groups <nc>the plurality</nc> of <nc>columns</nc> within <nc>the spreadsheet</nc>; and in <nc>response</nc> to detecting <nc>a selection</nc> of <nc>the preview thumbnail image</nc> corresponding to <nc>the first set</nc> of <nc>data cells</nc>, navigating to <nc>a display</nc> of <nc>the first set</nc> of <nc>data cells</nc> in <nc>a second graphical user interface content-view</nc> of <nc>the spreadsheet</nc>.
| 10
|
10. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the method</nc> further comprises: hiding <nc>a second set</nc> of <nc>data cells</nc> in <nc>the content-view</nc> of <nc>the spreadsheet</nc> based upon <nc>account permissions</nc> for <nc>access</nc> to <nc>the spreadsheet</nc>.
|
9548951
|
14145017
|
1. <nc>A method</nc>, comprising: identifying <nc>a message</nc> <nc>that</nc> is associated with <nc>two or more users</nc>, wherein <nc>the users</nc> include <nc>a sender</nc> and <nc>one or more recipients</nc>, and wherein <nc>the message</nc> includes <nc>one or more terms</nc>; determining <nc>a vague term</nc> of <nc>the terms</nc>, wherein <nc>the vague term</nc> includes <nc>a plurality</nc> of <nc>consecutive words</nc>, <nc>the consecutive words</nc> including <nc>at least a word</nc> and <nc>an additional word</nc>; determining that <nc>at least the word</nc> of <nc>the vague term</nc> is <nc>a reference</nc> to <nc>a given user</nc> of <nc>the users</nc>, <nc>the given user</nc> being <nc>the sender</nc> or one of <nc>the recipients</nc>; identifying, based on determining that <nc>the word</nc> is <nc>a reference</nc> to <nc>the given user</nc>, <nc>a user-restricted database</nc> associated with <nc>the given user</nc>, wherein <nc>the user-restricted database</nc> includes <nc>content</nc> personal to <nc>the given user</nc>, and wherein <nc>access</nc> to <nc>the user-restricted database</nc> is limitable by <nc>the given user</nc>; determining, based on <nc>the user-restricted database</nc>, additional information <nc>that</nc> is related to <nc>the vague term</nc>, wherein <nc>the user-restricted database</nc> is used in determining <nc>the additional information</nc> <nc>that</nc> is related to <nc>the vague term</nc> based on <nc>the user-restricted database</nc> being associated with <nc>the given user</nc> and based on <nc>the given user</nc> being referenced by <nc>the word</nc> of <nc>the vague term</nc>; and providing <nc>the additional information</nc> to at least one of <nc>the users</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , wherein providing <nc>the additional information</nc> includes providing <nc>a notification</nc> to <nc>the sender</nc> and wherein <nc>the sender</nc> is <nc>the creator</nc> of <nc>the message</nc>.
|
8639680
|
13465235
|
1. <nc>A method</nc> comprising: generating, by <nc>a system</nc> of <nc>one or more computers</nc>, <nc>a first data structure</nc> based on <nc>a web page</nc>, where <nc>the first data structure</nc> includes <nc>nodes</nc> corresponding to <nc>text</nc> <nc>that</nc> will be visually displayed from <nc>the web page</nc> and <nc>nodes</nc> corresponding to <nc>text</nc> <nc>that</nc> will not be visually displayed from <nc>the web page</nc> when <nc>the web page</nc> is rendered for <nc>display</nc> by <nc>a client device</nc>; generating, by <nc>the system</nc> of <nc>one or more computers</nc>, <nc>a second data structure</nc> for <nc>the web page</nc> based on <nc>the first data structure</nc>, <nc>the second data structure</nc> including <nc>nodes</nc> <nc>that</nc> correspond to <nc>text</nc> <nc>that</nc> will be visually displayed when <nc>the web page</nc> is rendered for <nc>display</nc> by <nc>the client device</nc>; comparing, by <nc>the system</nc> of <nc>one or more computers</nc>, <nc>nodes</nc> corresponding to <nc>text</nc> of <nc>the first data structure</nc> with <nc>nodes</nc> corresponding to <nc>text</nc> of <nc>the second data structure</nc> to identify <nc>the nodes</nc> corresponding to <nc>text</nc> <nc>that</nc> will not be visually displayed from <nc>the web page</nc> when <nc>the web page</nc> is rendered for <nc>display</nc> by <nc>the client device</nc>; and generating, by <nc>the system</nc> of <nc>one or more computers</nc>, <nc>weighting factors</nc> for <nc>the nodes</nc> corresponding to <nc>text</nc> <nc>that</nc> will not be visually displayed from <nc>the web page</nc> when <nc>the web page</nc> is rendered for <nc>display</nc> by <nc>the client device</nc> based on <nc>the nodes</nc> corresponding to <nc>text</nc> <nc>that</nc> will not be visually displayed from <nc>the web page</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: receiving, by <nc>the system</nc> of <nc>one or more computers</nc>, <nc>a search query</nc> from <nc>the client device</nc>, <nc>the search query</nc> including <nc>one or more query terms</nc>; identifying, by <nc>the system</nc> of <nc>one or more computers</nc>, multiple candidates of <nc>text</nc> responsive to <nc>the one or more query terms</nc> of <nc>the search query</nc>, where <nc>the multiple candidates</nc> of <nc>text</nc> includes <nc>text</nc> <nc>that</nc> will be visually displayed from <nc>the web page</nc> and <nc>text</nc> <nc>that</nc> will not be visually displayed from <nc>the web page</nc>; and selecting, by <nc>the system</nc> of <nc>one or more computers</nc>, <nc>text</nc> <nc>that</nc> is responsive to <nc>the one or more query terms</nc> of <nc>the search query</nc> from <nc>the multiple candidates</nc> of <nc>text</nc> based on <nc>the weighting factors</nc>.
|
9286548
|
14292294
|
1. <nc>A method</nc> for classifying <nc>search results</nc> by <nc>a search engine</nc> comprising <nc>an index</nc> of <nc>documents</nc> in <nc>a memory</nc> corresponding to <nc>a plurality</nc> of <nc>products</nc> offered by <nc>a plurality</nc> of <nc>online shopping sites</nc> and comprising <nc>text classifiers</nc> and <nc>image classifiers</nc>, <nc>a user interface</nc> enabling for enabling <nc>a user</nc> to browse <nc>products offerings</nc> organized according to <nc>the product taxonomy</nc> and receiving <nc>a search query</nc> from <nc>the user</nc> and providing <nc>the user</nc> with <nc>a set</nc> of <nc>search results</nc>, and <nc>a ranking engine</nc> for ranking <nc>the set</nc> of <nc>search results</nc> to enhance <nc>the user's experience</nc>, <nc>the method</nc> comprising: inferring <nc>a first distribution</nc> on <nc>a set</nc> of <nc>training data</nc> using <nc>a first classifier</nc>; inferring <nc>a second distribution</nc> on <nc>the set</nc> of <nc>training data</nc> using <nc>a second classifier</nc>; concatenating <nc>the first distribution</nc> and <nc>the second distribution</nc>; classifying <nc>each search result</nc> among <nc>a set</nc> of <nc>search results</nc> based on <nc>the concatenation</nc>; and ranking <nc>the search results</nc> based at least in <nc>part</nc> on <nc>the classification</nc>.
| 9
|
9. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining that classifying <nc>the search results</nc> will not be beneficial; and disregarding <nc>the ranking</nc> based on <nc>the classification</nc>.
|
8086960
|
11809612
|
1. <nc>A method</nc>, comprising: receiving <nc>an input</nc> to <nc>a body</nc> of <nc>an electronic markup language document</nc> <nc>that</nc> is being displayed in <nc>a window</nc>; and in <nc>response</nc> to <nc>said receiving</nc>: generating <nc>a comment</nc> based on or including <nc>the input</nc>; storing <nc>the comment</nc> in <nc>a comment section</nc> of <nc>a data structure</nc> for <nc>the electronic document</nc>, wherein said storing <nc>comprises</nc> formatting <nc>the comment</nc> using <nc>a tag</nc> to identify <nc>the comment</nc>, wherein <nc>the comment section</nc> of <nc>the data structure</nc> is separate from <nc>a body section</nc> of <nc>the data structure</nc>, wherein <nc>the body section</nc> of <nc>the data structure</nc> includes <nc>content</nc> for <nc>the body</nc> of <nc>the electronic markup language document</nc>; and displaying <nc>the body</nc> of <nc>the electronic markup language document</nc> and <nc>an action user interface element</nc> in <nc>the window</nc>, wherein <nc>the action user interface element</nc> can be activated to perform <nc>an action</nc> in <nc>regard</nc> to <nc>at least part</nc> of <nc>the comment</nc> stored in <nc>the comment section</nc> of <nc>the data structure</nc>, wherein <nc>the action</nc> includes one or more of <nc>view</nc>, accept, reject, modify, delete, insert, or replace.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the tag</nc> indicates <nc>an edit</nc> to <nc>the electronic document</nc>, <nc>the tag</nc> being stored in <nc>a pre-defined format</nc> within <nc>the comment section</nc> of <nc>the electronic document</nc>.
|
7483894
|
11751802
|
1. <nc>A method</nc> of searching for <nc>records</nc>, comprising: receiving <nc>a first query</nc> from <nc>a user</nc>; deriving <nc>an industry sector</nc> from <nc>the first query</nc>; using <nc>sector information</nc> to determine <nc>additional information</nc> to be used in deriving <nc>a second query</nc>, wherein <nc>the additional information</nc> is other than <nc>a hierarchical classification</nc>; requesting <nc>the additional information</nc> from <nc>the user</nc>; and deriving <nc>the second query</nc> from <nc>at least a portion</nc> of <nc>the first query</nc> and <nc>at least a portion</nc> of <nc>the additional information</nc> wherein <nc>the second query</nc> includes <nc>at least one section</nc> parameterized by <nc>an (attribute, value) pair</nc> used to direct at least two of <nc>the following actions</nc>: searching, breaking <nc>a tie</nc> in <nc>ranking results</nc>, describing <nc>relationships</nc> among <nc>sections</nc> of <nc>the second query</nc>, and expressing <nc>confidence measures</nc>.
| 15
|
15. <nc>The method</nc> of <nc>claim</nc> 1 , wherein at least one of <nc>the records</nc> corresponds to <nc>a physical product</nc>.
|
9177022
|
13287717
|
1. <nc>A method</nc> for <nc>processing queries</nc>, comprising: receiving <nc>user input</nc> creating <nc>a pipeline configuration</nc> for executing <nc>a query</nc> from <nc>an input device</nc>; receiving <nc>a first query</nc>; obtaining <nc>a context</nc> and <nc>conditions</nc> of <nc>the first query</nc>; determining <nc>rules</nc> based on <nc>the context</nc> and <nc>the conditions</nc> of <nc>the first query</nc> by utilizing <nc>the pipeline configuration</nc>, wherein <nc>the rules</nc> are triggered in <nc>response</nc> to receiving <nc>the first query</nc>; applying <nc>the rules</nc> to <nc>the first query</nc> to determine <nc>additional queries</nc> to execute; executing <nc>the additional queries</nc>; receiving <nc>supplemental results</nc> from <nc>execution</nc> of <nc>each</nc> of <nc>the additional queries</nc>; receiving <nc>core results</nc> from <nc>execution</nc> of <nc>the first query</nc>; and mixing <nc>the received supplemental results</nc> with <nc>the core results</nc> to form <nc>mixed results</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein executing <nc>the additional queries</nc> comprises: performing <nc>a federated search</nc> and executing <nc>parallel queries</nc> <nc>that</nc> are created in <nc>response</nc> to applying <nc>the rules</nc> to <nc>the first query</nc>.
|
8943481
|
13457285
|
1. <nc>A method</nc> for <nc>extensibility</nc> of <nc>binding definitions</nc> for <nc>a user interface application</nc>, <nc>the method</nc> comprising: performing, by <nc>a computer system</nc> programmed with <nc>code</nc> stored in <nc>a memory</nc> and executing by <nc>a processor</nc> of <nc>the computer system</nc> to configure <nc>the computer system</nc> into <nc>a machine</nc>: obtaining <nc>a framework</nc> having <nc>definitions</nc> of <nc>a first set</nc> of <nc>rules</nc> for <nc>a first grammar level</nc> used for <nc>interpretation</nc> of <nc>binding specifications</nc> to <nc>a user interface application</nc>, wherein <nc>the user interface application</nc> is incompatible with <nc>the first grammar level</nc>; performing <nc>a first transformation</nc> of <nc>the framework</nc> to generate <nc>the first set</nc> of <nc>rules</nc> for <nc>interpretation</nc> of <nc>the binding specifications</nc> in <nc>the first grammar level</nc>; performing <nc>a second transformation</nc> of <nc>the framework</nc> to generate <nc>a first presentation style</nc> for <nc>the first grammar level</nc>; obtaining <nc>the binding specifications</nc> in <nc>the first grammar level</nc>, <nc>the binding specification</nc> conforming to <nc>the first set</nc> of <nc>rules</nc>; and applying <nc>the first set</nc> of <nc>rules</nc> and <nc>the first presentation style</nc> to <nc>the binding specification</nc> to generate <nc>output binding specifications</nc> in <nc>a second grammar level</nc> compatible with <nc>the user interface application</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the first grammar level</nc> of <nc>the binding specifications</nc> is extensible.
|
8638299
|
13865553
|
1. <nc>A method</nc> of enabling <nc>input</nc> on <nc>an electronic device</nc> <nc>that</nc> comprises <nc>a set</nc> of <nc>keys</nc>, <nc>wherein each key</nc> in <nc>the set</nc> of <nc>keys</nc> has <nc>at least an alphabetic character</nc> and <nc>a non-alphabetic character</nc> assigned thereto, the method comprising: receiving <nc>an input sequence</nc> comprising <nc>one or more key selections</nc> of <nc>the set</nc> of <nc>keys</nc>; outputting, using <nc>an output apparatus</nc>, <nc>at least one interpretation</nc> of <nc>the input sequence</nc> <nc>that</nc> includes <nc>a non-alphabetic character</nc>, wherein <nc>the at least one interpretation</nc> is <nc>output</nc> at <nc>a position</nc> of <nc>relative priority</nc> based, at least in <nc>part</nc>, on <nc>a length</nc> of <nc>the input sequence</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining that <nc>the input sequence</nc> satisfies <nc>a predetermined condition</nc>, wherein <nc>the predetermined condition</nc> is satisfied when the one or <nc>more selected keys</nc> <nc>each</nc> has <nc>a numeric character</nc> and <nc>an alphabetic character</nc> assigned thereto and <nc>a quantity</nc> of <nc>key selections</nc> in <nc>the input sequence</nc> meets or exceeds <nc>a predetermined number</nc>, and wherein <nc>the at least one interpretation</nc> is <nc>a numeric interpretation</nc>.
|
9471551
|
13410921
|
1. <nc>A computer-implemented method</nc> comprising: Receiving <nc>various requests</nc> for <nc>content items</nc> to be displayed in <nc>various content item slots</nc> on <nc>various web resources</nc> presented at <nc>various user devices</nc>, wherein <nc>the various requests</nc> include <nc>information</nc> related to <nc>an amount</nc> of <nc>space</nc> available in <nc>the various content item slots</nc>; and for <nc>at least some</nc> of <nc>the various requests</nc>: identifying, using <nc>one or more processors</nc> and in <nc>response</nc> to <nc>the request</nc>, <nc>a content item</nc> from <nc>a set</nc> of <nc>eligible content items</nc> <nc>that</nc> are responsive to <nc>the request</nc>, <nc>the content item</nc> including <nc>a title portion</nc> and <nc>a body portion</nc>, <nc>the title portion</nc> including <nc>original text</nc> and being distinct from <nc>the body portion</nc>, wherein <nc>the body portion</nc> includes <nc>a plurality</nc> of <nc>lines</nc> of <nc>text</nc> including <nc>a first line</nc> of <nc>text</nc> and <nc>a second line</nc> of <nc>text</nc>; determining, based on <nc>the amount</nc> of <nc>space</nc> available in <nc>a content item slot</nc>, that <nc>the identified content item</nc> is too long to fit in <nc>the content item slot</nc>; and in <nc>response</nc> to determining that <nc>the identified content item</nc> is too long to fit in <nc>the content item slot</nc>: evaluating <nc>the body portion</nc> including determining when <nc>the body portion</nc> includes <nc>a complete phrase</nc> <nc>that</nc> is included in <nc>the plurality</nc> of <nc>lines</nc>, <nc>the evaluating</nc> including applying <nc>a test</nc> to <nc>one or more words</nc> in <nc>the body portion</nc>, wherein <nc>the evaluation</nc> is based, at least in <nc>part</nc>, on <nc>a size</nc> of <nc>the complete phrase</nc> and <nc>the amount</nc> of <nc>horizontal space</nc> in <nc>the title portion</nc>, as specified in <nc>the request</nc>; dynamically creating <nc>a modified content item</nc> from <nc>the content item</nc>, including promoting, using <nc>the one or more processors</nc>, <nc>the complete phrase</nc> into <nc>the title portion</nc> of <nc>the modified content item</nc>, wherein <nc>the title portion</nc> of <nc>the modified content item</nc> includes <nc>both the original text</nc> and <nc>the complete phrase</nc>; and providing <nc>the modified content item</nc> for <nc>presentation</nc> in <nc>the content item slot</nc> and in <nc>response</nc> to <nc>the content item request</nc>.
| 18
|
18. <nc>The method</nc> of <nc>claim</nc> 1 wherein promoting <nc>the complete phrase</nc> includes constructing <nc>a reference</nc> for <nc>the title portion</nc> after <nc>promotion</nc>.
|
9043290
|
13740529
|
1. A computer program product for rewriting <nc>a relational expression</nc>, <nc>the computer program product</nc> comprising: <nc>one or more computer-readable tangible storage devices</nc> and <nc>program instructions</nc> stored on at least one of <nc>the one or more storage devices</nc>, the program instructions comprising: <nc>program instructions</nc> to determine <nc>a first type</nc> of <nc>relational expression</nc>, wherein <nc>the first type</nc> of <nc>relational expression</nc> includes: a signed <nc>two's complement byte-length</nc>, a floating-point type, <nc>a packed decimal format</nc>, and <nc>a zoned decimal format</nc>; <nc>program instructions</nc> to determine <nc>a first segment</nc> of <nc>bytes</nc> and <nc>a second segment</nc> of <nc>bytes</nc>, wherein <nc>the first segment</nc> of <nc>bytes</nc> comprises <nc>leading bytes</nc> and <nc>the second segment</nc> of <nc>bytes</nc> comprises <nc>a last byte</nc> containing <nc>sign information</nc>; <nc>program instructions</nc> to generate <nc>a first set</nc> of <nc>conjunctive terms</nc>, wherein <nc>the first set</nc> of <nc>conjunctive terms</nc> are compared to <nc>the first segment</nc> of <nc>bytes</nc> and <nc>the second segment</nc> of <nc>bytes</nc>; and <nc>program instructions</nc> to rewrite <nc>the first type</nc> of <nc>the relational expression</nc> to <nc>a second type</nc> of <nc>relational expression</nc> based, at least in <nc>part</nc>, on <nc>the first set</nc> of <nc>conjunctive terms</nc>.
| 6
|
6. <nc>The computer program product</nc> of <nc>claim</nc> 1 , wherein <nc>the second type</nc> of <nc>relational expression</nc> comprises <nc>an unsigned binary relational operator</nc>.
|
8495499
|
13088307
|
1. <nc>A method</nc> comprising: receiving <nc>a first search query</nc> <nc>that</nc> was entered by <nc>a user</nc> and includes <nc>one or more terms</nc> <nc>that</nc> define <nc>the first search query</nc>; performing <nc>a search</nc> using <nc>the first search query</nc> to identify <nc>search results</nc> <nc>that</nc> are responsive to <nc>the first search query</nc>; based on performing <nc>the search</nc> using <nc>the first search query</nc>, identifying <nc>a first list</nc> of <nc>search results</nc> <nc>that</nc> are responsive to <nc>the first search query</nc>, the first list of <nc>search results</nc> including <nc>at least a first search result</nc> <nc>that</nc> is responsive to <nc>the first search query</nc> and that links to <nc>first electronic content</nc>; causing <nc>display</nc> of <nc>the first list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the first search query</nc>, <nc>the display</nc> of <nc>the first list</nc> of <nc>search results</nc> including <nc>the first search result</nc>; after causing <nc>display</nc> of <nc>the first list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the first search query</nc>, receiving <nc>user input</nc> selecting <nc>the first search result</nc> included in <nc>the display</nc> of <nc>the first list</nc> of <nc>search results</nc>; receiving <nc>a second search query</nc> <nc>that</nc> was entered by <nc>the user</nc> and includes <nc>one or more terms</nc> <nc>that</nc> define <nc>the second search query</nc>, <nc>the second search query</nc> being different than <nc>the first search query</nc>; performing <nc>a search</nc> using <nc>the second search query</nc> to identify <nc>search results</nc> <nc>that</nc> are responsive to <nc>the second search query</nc>; based on performing <nc>the search</nc> using <nc>the second search query</nc>, identifying <nc>a second list</nc> of <nc>search results</nc> <nc>that</nc> are responsive to <nc>the second search query</nc>, <nc>the second list</nc> of <nc>search results</nc> including <nc>at least a second search result</nc> <nc>that</nc> is responsive to <nc>the second search query</nc> and that <nc>links</nc> to <nc>second electronic content</nc>, <nc>the second list</nc> of <nc>search results</nc> being different than <nc>the first list</nc> of <nc>search results</nc> and <nc>the second search result</nc> being different than <nc>the first search result</nc>; causing <nc>display</nc> of <nc>the second list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the second search query</nc>, <nc>the display</nc> of <nc>the second list</nc> of <nc>search results</nc> including <nc>the second search result</nc>; after causing <nc>display</nc> of <nc>the second list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the second search query</nc>, receiving <nc>user input</nc> selecting <nc>the second search result</nc> included in <nc>the display</nc> of <nc>the second list</nc> of <nc>search results</nc>; subsequent to receiving <nc>user input</nc> selecting <nc>the first search result</nc> included in <nc>the display</nc> of <nc>the first list</nc> of <nc>search results</nc> and subsequent to receiving <nc>user input</nc> selecting <nc>the second search result</nc> included in <nc>the display</nc> of <nc>the second list</nc> of <nc>search results</nc>, receiving <nc>a third search query</nc> <nc>that</nc> was entered by <nc>the user</nc> and includes <nc>one or more terms</nc> <nc>that</nc> define <nc>the third search query</nc>, <nc>the third search query</nc> being different than <nc>the first search query</nc> and being different than <nc>the second search query</nc>; performing <nc>a search</nc> using <nc>the third search query</nc> to identify <nc>search results</nc> <nc>that</nc> are responsive to <nc>the third search query</nc>; based on performing <nc>the search</nc> using <nc>the third search query</nc>, identifying <nc>a third list</nc> of <nc>search results</nc> <nc>that</nc> are responsive to <nc>the third search query</nc>, <nc>the third list</nc> of <nc>search results</nc> being different than <nc>the first list</nc> of <nc>search results</nc> and being different than <nc>the second list</nc> of <nc>search results</nc>, <nc>the third list</nc> of <nc>search results</nc> including <nc>the first search result</nc> that <nc>links</nc> to <nc>the first electronic content</nc>, the second search result that links to <nc>the second electronic content</nc>, and a third search result that links to <nc>third electronic content</nc>, <nc>that</nc> has not been selected by <nc>the user</nc> prior to receiving <nc>the third search query</nc>, and <nc>that</nc> is different than <nc>the first search result</nc> and <nc>the second search result</nc>; based on <nc>the selection</nc>, prior to receiving <nc>the third search query</nc>, of <nc>the first search result</nc> included in <nc>the display</nc> of <nc>the first list</nc> of <nc>search results</nc>, <nc>the selection</nc>, prior to receiving <nc>the third search query</nc>, of <nc>the second search result</nc> included in <nc>the display</nc> of <nc>the second list</nc> of <nc>search results</nc>, and <nc>the third list</nc> of <nc>search results</nc> including <nc>the first search result</nc> and <nc>the second search result</nc>, <nc>grouping</nc>, by <nc>at least one processor</nc> and within <nc>a display</nc> of <nc>the third list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the third search query</nc>, <nc>a first representation</nc> of <nc>the first search result</nc> and <nc>a second representation</nc> of <nc>the second search result</nc> together in <nc>a group</nc> separate from <nc>a third representation</nc> of <nc>the third search result</nc> <nc>that</nc> has not been selected by <nc>the user</nc>, even though <nc>relevancy ratings</nc>, to <nc>the third search query</nc>, for <nc>the first search result</nc>, the second search result, and <nc>the third search result</nc> do not suggest grouping <nc>the first search result</nc> and <nc>the second search result</nc> together in <nc>the group</nc> separate from <nc>the third search result</nc> <nc>that</nc> has not been selected by <nc>the user</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising, based on <nc>the selection</nc>, prior to receiving <nc>the third search query</nc>, of <nc>the first search result</nc> included in <nc>the display</nc> of <nc>the first list</nc> of <nc>search results</nc>, <nc>the selection</nc>, prior to receiving <nc>the third search query</nc>, of <nc>the second search result</nc> included in <nc>the display</nc> of <nc>the second list</nc> of <nc>search results</nc>, and <nc>the third list</nc> of <nc>search results</nc> including <nc>the first search result</nc> and <nc>the second search result</nc>, causing <nc>display</nc>, within <nc>the display</nc> of <nc>the third list</nc> of <nc>search results</nc> identified based on performing <nc>the search</nc> using <nc>the third search query</nc>, of <nc>a first graphical indicator</nc> <nc>that</nc> is proximate to <nc>the first representation</nc> of <nc>the first search result</nc>, <nc>that</nc> distinguishes <nc>the first representation</nc> of <nc>the first search result</nc> from <nc>the third representation</nc> of <nc>the third search result</nc>, and <nc>that</nc> indicates that <nc>the first search result</nc> was previously selected, and a second graphical indicator <nc>that</nc> is proximate to <nc>the second representation</nc> of <nc>the second search result</nc>, <nc>that</nc> distinguishes <nc>the second representation</nc> of <nc>the second search result</nc> from <nc>the third representation</nc> of <nc>the third search result</nc>, and <nc>that</nc> indicates that <nc>the second search result</nc> was previously selected.
|
9424315
|
12099076
|
1. <nc>A method</nc> for <nc>scheduling database tasks</nc> for <nc>a relational query</nc>, said <nc>method</nc> comprising: receiving <nc>at least a portion</nc> of <nc>a query execution plan</nc> comprising <nc>a set</nc> of <nc>fragments</nc> with <nc>respective tasks</nc> having <nc>corresponding database virtual address</nc> ranges; mapping <nc>the database virtual address</nc> ranges to <nc>a set</nc> of <nc>memory pages</nc>; requesting <nc>respective locks</nc> for <nc>the set</nc> of <nc>memory pages</nc>; locking <nc>memory pages</nc> for <nc>the selected tasks</nc>; marking one or more of <nc>the tasks</nc> as being ready for <nc>execution</nc> based on whether <nc>locks</nc> have been granted for <nc>all</nc> of <nc>the memory page requests</nc>; and scheduling <nc>one or more tasks</nc> for <nc>execution</nc> in <nc>a dataflow architecture hardware accelerator</nc> coupled to <nc>the memory pages</nc> when <nc>locks</nc> have been obtained for <nc>all</nc> of <nc>its memory pages</nc>, <nc>wherein execution</nc> of <nc>the one or more tasks</nc> is based on <nc>execution</nc> of <nc>machine code instructions</nc> formatted for <nc>the dataflow architecture hardware accelerator</nc> and specifying <nc>a dataflow</nc> of <nc>data</nc> through <nc>the dataflow architecture hardware accelerator</nc>.
| 15
|
15. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining <nc>a deadlock</nc> exists when there is <nc>no running task</nc> and <nc>no task</nc> can be dispatched; and resolving <nc>the deadlock</nc> based on rescheduling <nc>all uncompleted tasks</nc>.
|
9177069
|
13111459
|
1. <nc>A method</nc> comprising: determining <nc>a target geographic feature</nc> <nc>that</nc> has <nc>insufficient targeting information</nc> associated therewith, wherein <nc>the insufficient targeting information</nc> is insufficient to provide <nc>targeted content</nc> associated with <nc>the target geographic feature</nc>, <nc>the target geographic feature</nc> defining <nc>a location</nc>; determining <nc>one or more similar geographic features</nc> to <nc>the target geographic feature</nc>, <nc>each similar geographic feature</nc> including targeting <nc>information</nc> and defining <nc>a different non-overlapping location</nc> from <nc>the target geographic location</nc>, wherein <nc>the one or more similar geographic features</nc> are determined based at least in <nc>part</nc> by identifying <nc>a geographic feature</nc> having <nc>one or more excess queries</nc> in common with <nc>the target geographic feature</nc>, wherein <nc>each excess query</nc> is <nc>a query</nc> associated with and exceeds <nc>an expected query</nc> count for <nc>each</nc> of <nc>the one or more similar geographic features</nc> and <nc>the target geographic feature</nc>; attributing targeting <nc>information</nc> associated with at least one of <nc>the one or more similar geographic features</nc> to <nc>the target geographic feature</nc>; and serving <nc>content</nc> responsive to <nc>queries</nc> <nc>that</nc> relate to <nc>the target geographic feature</nc> based at least in <nc>part</nc> on <nc>the attributed targeting information</nc>; wherein for <nc>each</nc> of <nc>the target geographic feature</nc> and <nc>the one or more similar geographic features</nc>, <nc>the expected query</nc> count associated with <nc>the query</nc> is computed as <nc>a function</nc> of <nc>a total number</nc> of <nc>queries</nc> received over <nc>a time period</nc> for <nc>the target geographic feature</nc> or <nc>the one or more similar geographic features</nc> and <nc>a query share</nc> associated with <nc>the query</nc>; wherein <nc>the query share</nc> is computed as <nc>a ratio</nc> of <nc>the number</nc> of <nc>times</nc> that <nc>the query</nc> was received from <nc>user devices</nc> in <nc>a baseline geographic region</nc> relative to <nc>a total number</nc> of <nc>queries</nc> <nc>that</nc> have been received from <nc>user devices</nc> in <nc>the baseline geographic region</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 where <nc>serving content</nc> comprises serving <nc>an advertisement</nc>.
|
9104969
|
14698936
|
1. <nc>A system</nc> configured to utilize <nc>semantic analysis</nc> to set <nc>a processing level</nc> for <nc>processing measurements</nc> of <nc>affective response</nc>, comprising: <nc>a semantic analyzer</nc> configured to: receive <nc>a first segment</nc> of <nc>content</nc>, analyze <nc>the first segment</nc> utilizing <nc>semantic analysis</nc>, and output <nc>a first indication</nc> that <nc>a first value</nc> related to <nc>a predicted emotional response</nc> to <nc>the first segment</nc> does not reach <nc>a first predetermined threshold</nc>; wherein <nc>the first segment</nc> comprises <nc>data</nc> representing <nc>text</nc>; and <nc>a hardware-based controller</nc> configured to select, based on <nc>the first indication</nc>, <nc>a first processing level</nc> for <nc>a processor</nc> to process <nc>measurements</nc> of <nc>affective response</nc> of <nc>a user</nc> to <nc>the first segment</nc>; wherein <nc>the semantic analyzer</nc> is further configured to: receive <nc>a second segment</nc> of <nc>content</nc>, analyze <nc>the second segment</nc> utilizing <nc>semantic analysis</nc>, and output <nc>a second indication</nc> that <nc>a second value</nc> related to <nc>a predicted emotional response</nc> to <nc>the second segment</nc> reaches <nc>a second predetermined threshold</nc>; wherein <nc>the second segment</nc> comprises <nc>data representing text</nc>; wherein <nc>the hardware-based controller</nc> is further configured to select, based on <nc>the second indication</nc>, <nc>a second processing level</nc> for <nc>a processor</nc> to process <nc>measurements</nc> of <nc>affective response</nc> of <nc>the user</nc> to <nc>the second segment</nc>; wherein, per <nc>volume unit</nc> of <nc>measurement data</nc>, <nc>the number</nc> of <nc>computation cycles</nc> utilized by <nc>the processor</nc> to process, at <nc>the first processing level</nc>, <nc>the measurements</nc> of <nc>the affective response</nc> of <nc>the user</nc> to <nc>the first segment</nc>, is at least 50% lower than <nc>the number</nc> of <nc>computation cycles</nc> utilized by <nc>the processor</nc> to process, at <nc>the second processing level</nc>, <nc>the measurements</nc> of <nc>the affective response</nc> of <nc>the user</nc> to <nc>the second segment</nc>.
| 9
|
9. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the semantic analyzer</nc> applies <nc>Latent Semantic Analysis</nc> in <nc>order</nc> to associate <nc>a segment</nc> of <nc>content</nc> with <nc>a likely subject</nc>.
|
8046228
|
11349567
|
1. <nc>A speech user agent</nc>, comprising: <nc>an output</nc> for outputting <nc>a uniform resource location</nc> (<nc>URL) directive</nc>; <nc>an input</nc> for receiving <nc>grammar HTML</nc>; and <nc>an output</nc> for outputting <nc>a URL</nc> with <nc>arguments</nc>.
| 2
|
2. <nc>The speech user agent</nc> of <nc>claim</nc> 1 , wherein said <nc>output</nc> for outputting a uniform resource location (<nc>URL</nc>) directive is coupled to <nc>an input</nc> of <nc>a web browser</nc>.
|
7933762
|
10826630
|
1. <nc>A machine-based method</nc> comprising: in <nc>connection</nc> with <nc>a project</nc> in <nc>which</nc> <nc>a predictive model</nc> is generated based on <nc>historical data</nc> about <nc>a system</nc> being modeled <nc>selecting variables</nc> having <nc>at least a first predetermined level</nc> of <nc>significance</nc> from <nc>a pool</nc> of <nc>potential predictor variables</nc> associated with <nc>the historical data</nc>, to form <nc>a first population</nc> of <nc>predictor variables</nc>, extending <nc>the first population</nc> of <nc>predictor variables</nc> to include <nc>cross products</nc> of <nc>at least two variables</nc>, <nc>each</nc> being from <nc>the first population</nc> of <nc>predictor variables</nc>, selecting <nc>variables</nc> having <nc>at least a second predetermined level</nc> of <nc>significance</nc> from <nc>the extended first population</nc> of <nc>predictor variables</nc> to form <nc>a second population</nc> of <nc>predictor variables</nc>, extending <nc>the second population</nc> of <nc>predictor variables</nc> to include <nc>cross products</nc> of <nc>at least two variables</nc>, at least one of <nc>the variables</nc> for at least one of <nc>the cross products</nc> being from <nc>the pool</nc> of <nc>potential predictor variables</nc> <nc>that</nc> are associated with <nc>the historical data</nc> and having less than <nc>the first predetermined level</nc> of <nc>significance</nc>, selecting <nc>variables</nc> having <nc>at least a third predetermined level</nc> of <nc>significance</nc> from <nc>the extended second population</nc> of <nc>predictor variables</nc> to form <nc>a third population</nc> of <nc>predictor variables</nc>, automatically selecting <nc>a model generation method</nc> from <nc>a set</nc> of <nc>available model generation methods</nc> to match <nc>characteristics</nc> of <nc>the historical data</nc>, generating <nc>a possible model</nc> of <nc>the third population</nc> of <nc>predictor variables</nc> using <nc>a subsample</nc> of <nc>the historical data</nc> by <nc>the model generation method</nc>, determining whether <nc>the possible model</nc> generalizes to <nc>the historical data</nc> other than <nc>the subsample</nc>, applying <nc>the possible model</nc> to <nc>all</nc> of <nc>the historical data</nc> to generate <nc>a final model</nc>, cross-<nc>validating the final model</nc> using <nc>random portions</nc> of <nc>the historical data</nc>, and interacting with <nc>the system</nc> being modeled based on <nc>the final model</nc>.
| 12
|
12. <nc>The method</nc> of <nc>claim</nc> 1 in <nc>which</nc> <nc>a user</nc> is enabled to select <nc>at least one validation dataset</nc> and invoke <nc>a model process validation method</nc>.
|
9225679
|
14171137
|
1. <nc>A system</nc> comprising: <nc>a server</nc> in <nc>communication</nc> with <nc>a vehicle computing system</nc> (<nc>VCS</nc>) via <nc>a transceiver</nc> and configured to: in <nc>response</nc> to <nc>an electronic-mail message</nc> identifying <nc>the VCS</nc>, transmit <nc>information</nc> from <nc>the message</nc> received at <nc>the server</nc> to <nc>an application</nc> on <nc>the VCS</nc>, <nc>the information</nc> pre-specified for <nc>utilization</nc> by <nc>the application</nc> and associated with <nc>a predefined category</nc> identified via <nc>a classification model</nc> classifying <nc>an element</nc> corresponding to <nc>the predefined category</nc> in <nc>text</nc> of <nc>the message</nc>.
| 9
|
9. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the pre-specified information</nc> is <nc>navigation information</nc> based on <nc>the application</nc> being <nc>a navigation application</nc>.
|
9075983
|
14003814
|
1. <nc>A method</nc> for authorizing <nc>access</nc>, comprising <nc>the step</nc> of: generating for <nc>display</nc> <nc>at least one distorted string</nc> of <nc>alphanumeric characters</nc>, in <nc>combination</nc> with at least one of <nc>a glyph</nc>, <nc>picture</nc> or <nc>symbol</nc>, <nc>the glyph</nc>, <nc>picture</nc> or <nc>symbol</nc> being foreign to <nc>a target audience</nc>, wherein <nc>the generating</nc> includes separating <nc>the at least one distorted string</nc> of <nc>alphanumeric characters</nc> into <nc>two or more strings</nc>, and adding <nc>the at least one glyph</nc>, <nc>picture</nc> or <nc>symbol</nc> to <nc>one or more ends</nc> of <nc>the two or more strings</nc> to form <nc>at least one string</nc> of <nc>random alphanumeric characters</nc> <nc>that</nc> includes at least one of <nc>the glyph</nc>, <nc>picture</nc> or <nc>symbol</nc>; and comparing <nc>a response</nc> of <nc>a user</nc> entered in <nc>reaction</nc> to <nc>the distorted string</nc> of <nc>alphanumeric characters</nc> to <nc>a reference string</nc> of <nc>characters</nc> to determine whether to grant <nc>access</nc>.
| 6
|
6. <nc>The method</nc> according <nc>to claim</nc> 1 wherein <nc>the at least one string</nc> of <nc>random alphanumeric characters</nc> comprises <nc>at least one word</nc>.
|
8180629
|
12170433
|
1. <nc>A computer</nc> implemented <nc>method</nc> of generating <nc>declared patterns</nc> from <nc>components</nc> of <nc>a sentence</nc>, wherein <nc>each</nc> of said <nc>declared patterns</nc> is <nc>a sequence</nc> of <nc>pattern units</nc>, wherein <nc>each</nc> of <nc>said pattern units</nc> corresponds to <nc>a portion</nc> of <nc>text</nc> in <nc>said sentence</nc>, comprising: providing <nc>a computer-readable memory</nc> tangibly embodying <nc>said method</nc> as <nc>a computer program</nc>, <nc>wherein execution</nc> of <nc>the computer program</nc> in <nc>a computer</nc> comprising <nc>one or more processors</nc> is configured to: providing <nc>a conceptionary</nc>, wherein said <nc>conceptionary</nc> is <nc>a knowledge representation</nc> of <nc>items</nc>, wherein said <nc>items</nc> represent <nc>concepts</nc>, <nc>instances</nc>, <nc>relationships</nc>, <nc>characteristics</nc>, <nc>values</nc>, <nc>units</nc> of <nc>measure</nc>, and <nc>sets</nc> and <nc>aggregations</nc> of <nc>the items</nc>; providing <nc>a parts</nc> of <nc>speech tagger database</nc>; providing <nc>a first dictionary</nc> comprising <nc>dictionary entries</nc> of <nc>words</nc> and <nc>phrases</nc>, wherein said <nc>first dictionary</nc> further comprises declared categories and <nc>emergent categories</nc>, further wherein each of said <nc>dictionary entries</nc> is <nc>a term</nc> composed of <nc>single words</nc> and <nc>phrases</nc> and specifies <nc>individual senses</nc> of <nc>said term</nc>; providing <nc>a database</nc> of <nc>equivalent pattern specification sets</nc> for <nc>a given language</nc>, wherein said <nc>equivalent pattern specification sets</nc> represent <nc>different ways</nc> of saying <nc>the same thing</nc> in said given <nc>language</nc>; providing <nc>a second dictionary</nc> in <nc>a target language</nc> comprising <nc>an equivalent name</nc> set for <nc>each sense</nc> of <nc>a word</nc>, wherein said <nc>equivalent name set</nc> is <nc>a set</nc> of <nc>terms</nc> <nc>that</nc> are semantically close in <nc>meaning</nc>; for <nc>each sentence</nc> in <nc>a document corpus</nc>; tagging <nc>parts</nc> of <nc>speech</nc> in <nc>the sentence</nc> for identifying <nc>parts</nc> of <nc>speech</nc> of <nc>each word</nc> and <nc>phrase</nc> in <nc>the sentence</nc>; chunking <nc>the sentence</nc> using said <nc>identified parts</nc> of <nc>speech</nc> of <nc>each word</nc> and <nc>phrase</nc> to generate <nc>pattern units</nc>; identifying <nc>grammatical roles</nc> and <nc>senses</nc> of <nc>said generated pattern units</nc> by applying <nc>said first dictionary</nc> and said <nc>database</nc> of <nc>equivalent pattern specification sets</nc>; identifying <nc>an equivalent name</nc> set for <nc>each</nc> of <nc>the generated pattern units</nc> by applying said <nc>second dictionary</nc> and <nc>the conceptionary</nc>; and generating <nc>the declared patterns</nc> for <nc>the sentence</nc> using said <nc>identified equivalent name</nc> set for <nc>each</nc> of <nc>the generated pattern units</nc>.
| 6
|
6. <nc>The computer</nc> implemented <nc>method</nc> of <nc>claim</nc> 1 , further comprising <nc>a step</nc> of constructing <nc>an equivalent name</nc> set for <nc>a term</nc> in <nc>absence</nc> of <nc>said equivalent name</nc> set for <nc>said term</nc> in said <nc>second dictionary</nc>, wherein said <nc>construction</nc> of <nc>the equivalent name</nc> <nc>set</nc> of <nc>the term</nc> comprises <nc>the steps</nc> of: providing <nc>an extended dictionary</nc> with <nc>senses</nc>, wherein <nc>each</nc> of said <nc>senses</nc> has <nc>synonyms</nc>; looking up <nc>dictionary entry</nc> for <nc>the term</nc> in <nc>said extended dictionary</nc> with <nc>the senses</nc> and identifying <nc>dictionary entries</nc> corresponding to said <nc>synonyms</nc> of said senses of <nc>the term</nc>; identifying <nc>senses</nc> of <nc>each</nc> of <nc>the synonyms</nc> from <nc>the extended dictionary</nc> with <nc>the senses</nc>; and adding said <nc>identified senses</nc> to <nc>the equivalent name</nc> set of <nc>the senses</nc> of <nc>the term</nc> <nc>whose equivalent name set</nc> is being constructed, if <nc>identified senses</nc> are equal to <nc>the term</nc>.
|
7895241
|
11873366
|
1. <nc>A method</nc> for integrating <nc>oilfield data</nc> in <nc>a plurality</nc> of <nc>formats</nc>, <nc>the method</nc> comprising: storing <nc>the oilfield data</nc> associated with <nc>a plurality</nc> of <nc>oilfield entities</nc> in <nc>a first data repository</nc>; obtaining, using <nc>a computer</nc>, <nc>a first target metamodel</nc> comprising <nc>a first structural description</nc> of <nc>a first plurality</nc> of <nc>data entities</nc> of <nc>the first data repository</nc>, wherein <nc>the first structural description</nc> describes <nc>database rows</nc> and <nc>columns</nc> of <nc>the first data repository</nc>; obtaining, using <nc>the computer</nc>, <nc>a domain metamodel</nc> interleaved with <nc>a first mapping specification</nc>, <nc>the domain metamodel</nc> comprising <nc>a second structural description</nc> of <nc>a domain model</nc>, defining <nc>a plurality</nc> of <nc>domain objects</nc> within <nc>a hierarchy</nc> of <nc>an oilfield domain</nc>, for representing <nc>the plurality</nc> of <nc>oilfield entities</nc> in <nc>an application programming interface</nc> (<nc>API</nc>); associating, using <nc>the computer</nc>, <nc>the first plurality</nc> of <nc>data entities</nc> of <nc>the first target metamodel</nc> with <nc>the plurality</nc> of <nc>domain objects</nc> of <nc>the domain metamodel</nc> using <nc>the first mapping specification</nc>; forming, using <nc>the computer</nc>, <nc>the API</nc> based on <nc>the domain metamodel</nc>, <nc>the first target metamodel</nc>, and <nc>the first mapping specification</nc>; generating <nc>a domain model source code</nc> based on <nc>the domain metamodel</nc>, <nc>the first target metamodel</nc>, and the first mapping specification using <nc>an automatic code generator</nc>; compiling <nc>the domain model source code</nc> to implement <nc>the API</nc>; generating <nc>a plurality</nc> of <nc>domain object instances</nc> corresponding to <nc>the plurality</nc> of <nc>oilfield entities</nc>, wherein at least one of <nc>the plurality</nc> of <nc>domain object instances</nc> is instantiated from <nc>a domain object</nc> of <nc>the plurality</nc> of <nc>domain objects</nc> compiled from <nc>the domain model source code</nc>; and identifying <nc>an association</nc> between <nc>the plurality</nc> of <nc>oilfield entities</nc> with <nc>the first plurality</nc> of <nc>data entities</nc> based on the at least one of <nc>the plurality</nc> of <nc>domain object instances</nc>; receiving, by <nc>the API</nc>, <nc>a database operation</nc> for <nc>the domain metamodel</nc>, wherein <nc>the database operation</nc> comprises <nc>a plurality</nc> of <nc>elements</nc> <nc>that</nc> generate <nc>a result</nc> using <nc>data</nc> in <nc>the domain metamodel</nc>; determining that <nc>one element</nc> of <nc>the plurality</nc> of <nc>elements</nc> in <nc>the database operation</nc> for <nc>the domain metamodel</nc> is needed to generate <nc>the result</nc> in <nc>a form</nc> required by <nc>the database operation</nc>; executing <nc>the database operation</nc> after <nc>a status</nc> of <nc>the one element</nc> in <nc>the database operation</nc> is changed from optional to mandatory, and generating <nc>the result</nc> in <nc>the form</nc> required by <nc>the database operation</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , wherein forming <nc>the API comprises</nc>: forming <nc>an interface layer</nc> of <nc>the API</nc> based on <nc>the domain metamodel</nc>; and forming <nc>an implementation layer</nc> of <nc>the API</nc> based on <nc>the first target metamodel</nc>, <nc>the domain metamodel</nc>, and <nc>the first mapping specification</nc>.
|
8050503
|
11773820
|
1. <nc>A system</nc> for <nc>computer vision</nc>, <nc>the system</nc> including: <nc>a plurality</nc> of <nc>images</nc>; <nc>a signature processor</nc> of <nc>at least one processing device</nc> adapted to generate <nc>a signature</nc> based at least in <nc>part</nc> on <nc>a curvelet</nc> transform, <nc>the signature</nc> including <nc>a portion</nc> of <nc>a plurality</nc> of <nc>curvelet coefficients</nc> of <nc>the curvelet</nc> transform <nc>that</nc> is significant, wherein <nc>the signature</nc> is <nc>a vector</nc> of <nc>real numbers</nc> with a length less than <nc>a number</nc> of <nc>corresponding image pixels</nc>; and <nc>a matching processor</nc> of <nc>the at least one processing device</nc> adapted to receive <nc>a query image</nc>, wherein <nc>the matching processor</nc> is adapted to determine <nc>a query signature</nc> for <nc>the query image</nc> using <nc>the signature processor</nc>, and wherein <nc>the matching processor</nc> is adapted to determine <nc>at least one matching image</nc> from <nc>the plurality</nc> of <nc>images</nc> based at least in <nc>part</nc> on <nc>the query signature</nc>.
| 11
|
11. <nc>The system</nc> of <nc>claim</nc> 1 , wherein the determined at least on <nc>matching image</nc> is determined by ranking at least one of <nc>the plurality</nc> of <nc>images</nc>.
|
8190999
|
10850399
|
1. A method comprising: <nc>a computer</nc> creating <nc>a plurality</nc> of <nc>topics</nc>; the computer assigning <nc>a statement classification method</nc> to <nc>each</nc> of <nc>the plurality</nc> of <nc>topics</nc>; the computer submitting <nc>a plurality</nc> of <nc>instant messaging statements</nc> to <nc>an instant messaging channel</nc>; <nc>the computer</nc> associating <nc>each</nc> of <nc>the plurality</nc> of <nc>instant messaging statements</nc> with one of <nc>the plurality</nc> of <nc>topics</nc> in <nc>accordance</nc> with <nc>an assigned statement classification method</nc>; <nc>the computer</nc> grouping together <nc>each</nc> of <nc>the plurality</nc> of <nc>instant messaging statements</nc> having <nc>a common topic</nc> into <nc>a plurality</nc> of <nc>topic groups</nc>; <nc>the computer</nc> displaying <nc>each</nc> of <nc>the plurality</nc> of <nc>topic groups</nc> in <nc>a separate window</nc> on <nc>a computer display</nc>; wherein <nc>each</nc> of <nc>the topics</nc> comprises <nc>a topic name</nc>; wherein <nc>the topic name</nc> of <nc>each</nc> of <nc>the topics</nc> is negotiable, thereby allowing <nc>any participant</nc> to propose <nc>a change</nc> to <nc>the topic name</nc>; and wherein <nc>the change</nc> is subject to <nc>an approval</nc> by <nc>a group leader</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 further comprising: wherein <nc>the assigned statement classification method</nc> is by <nc>a keyword</nc>; the computer assigning <nc>the keyword</nc> to one of <nc>the topics</nc>; <nc>the computer</nc> determining whether <nc>each</nc> of <nc>the plurality</nc> of <nc>instant messaging statements</nc> contains <nc>the keyword</nc>; and <nc>the computer</nc> responsive to <nc>a determination</nc> that one of <nc>the plurality</nc> of <nc>instant messaging statements</nc> contains <nc>the keyword</nc>, associating <nc>an instant messaging statement</nc> containing <nc>the keyword</nc> with <nc>a topic</nc> associated with <nc>the keyword</nc>.
|
9275018
|
13558699
|
1. A computer-implemented method, comprising: intercepting, at <nc>a proxy server</nc> including <nc>one or more processors</nc>, <nc>a request</nc> for <nc>a source document</nc> representing <nc>a web page</nc>, <nc>the request</nc> being transmitted from <nc>a remote computing device</nc> to <nc>a remote web server</nc> via <nc>a network</nc>, <nc>the request</nc> including <nc>web browser information</nc> indicating <nc>web browser software</nc> executing on <nc>the remote computing device</nc>; intercepting, at <nc>the proxy server</nc>, the source document being transmitted from <nc>the remote web server</nc> to <nc>the remote computing device</nc> via <nc>the network</nc> in <nc>response</nc> to <nc>the request</nc>, <nc>the source document</nc> including <nc>a text</nc> and specifying <nc>one or more fonts</nc> in <nc>which</nc> to display <nc>the text</nc>; rendering, at <nc>the proxy server</nc>, <nc>the web page</nc> using <nc>the source document</nc>, <nc>the web page</nc> including <nc>the text</nc> displayed in <nc>the one or more fonts</nc>; determining, at <nc>the proxy server</nc>, <nc>unique characters</nc> displayed at <nc>the web page</nc> for <nc>each</nc> of <nc>the specified one or more fonts</nc> in <nc>which</nc> <nc>the text</nc> is displayed; obtaining, at <nc>the proxy server</nc>, <nc>one or more font subsets</nc> based on <nc>the unique characters</nc>, wherein at least one of <nc>the one or more font subsets</nc> includes <nc>the unique characters</nc> in <nc>the source document</nc> and <nc>one or more additional characters</nc> related to <nc>the unique characters</nc> in <nc>the source document</nc>, and wherein <nc>the one or more additional characters</nc> related to <nc>the unique characters</nc> include at least one of (<nc>i</nc>) <nc>one or more characters</nc> having <nc>a different case</nc> than one or more of <nc>the unique characters</nc> and <nc>(ii</nc>) one or more characters having <nc>a different accent</nc> than one or more of <nc>the unique characters</nc>; determining, at <nc>the proxy server</nc>, whether <nc>the one or more fonts</nc> in <nc>which</nc> <nc>the text</nc> is displayed form <nc>a font family</nc> having <nc>a plurality</nc> of <nc>fonts</nc>, <nc>each</nc> of <nc>the plurality</nc> of <nc>fonts</nc> having at least one of <nc>a different weight</nc> and <nc>a different style</nc>; determining, at <nc>the proxy server</nc>, whether <nc>the web browser software</nc> indicated by <nc>the web browser information</nc> is capable of displaying <nc>the font family</nc> having <nc>the plurality</nc> of <nc>fonts</nc>; and modifying, at <nc>the proxy server</nc>, <nc>the source document</nc> by embedding <nc>the plurality</nc> of <nc>fonts</nc> therein to obtain <nc>a modified source document</nc> when <nc>the one or more fonts</nc> displayed at <nc>the web page</nc> form <nc>the font family</nc> having <nc>the plurality</nc> of <nc>fonts</nc> and <nc>the web browser software</nc> indicated by <nc>the web browser information</nc> is incapable of displaying <nc>the font family</nc> having <nc>the plurality</nc> of <nc>fonts</nc>; and transmitting, from <nc>the proxy server</nc> to <nc>the remote computing device</nc>, <nc>information</nc> specifying <nc>the one or more font subsets</nc> when <nc>the web browser software</nc> indicated by <nc>the web browser information</nc> is capable of displaying <nc>the font family</nc> having <nc>the plurality</nc> of <nc>fonts</nc>, <nc>the modified source document</nc>, or <nc>a command</nc> and <nc>a single font</nc> subset.
| 3
|
3. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , further comprising in <nc>response</nc> to being unable to locate or identify <nc>one or more particular characters</nc>, obtaining <nc>the one or more font subsets</nc> having less than <nc>all</nc> of <nc>the unique characters</nc> in <nc>the source document</nc>.
|
8156429
|
11576820
|
1. <nc>A server</nc> communicatively coupled to <nc>an Internet Protocol “IP” network</nc> and configured to transfer <nc>markup language based content</nc> <nc>that</nc> represents <nc>a web page</nc> to <nc>a client computer</nc> within <nc>the network</nc>, <nc>the server</nc> comprising: <nc>a dispatcher module</nc> that<nc>: interfaces</nc> to <nc>a content source</nc> and receives <nc>a markup language content</nc> from <nc>the content source</nc>, wherein <nc>the received markup language content</nc> is targeted toward <nc>a browser application</nc> at <nc>a requesting client computer</nc> and includes <nc>one or more browser links</nc>; provides <nc>at least a portion</nc> of <nc>the received markup language content</nc> to <nc>a markup language object module</nc>; receives <nc>modified markup language content</nc> from <nc>the markup language object module</nc>; and sends <nc>the modified markup language content</nc> toward <nc>the browser application</nc> at <nc>the requesting client computer</nc>; and <nc>the markup language object module</nc> that: receives <nc>the provided markup language content</nc> from <nc>the dispatcher module</nc>; parses <nc>the provided markup language content</nc>; identifies <nc>a browser link</nc> within <nc>the provided markup language content</nc>; traverses <nc>the browser link</nc> to obtain <nc>the corresponding object</nc>; replaces <nc>the browser link</nc> in <nc>the provided markup language content</nc> with <nc>the corresponding object</nc> to modify <nc>the provided markup language content</nc>; and provides <nc>the modified markup language content</nc> to <nc>the dispatcher module</nc>; wherein <nc>the server</nc> operates in <nc>an intermediate node</nc> communicatively positioned between <nc>the content source</nc> and <nc>the requesting client</nc>; wherein <nc>the modified markup language content</nc> <nc>that</nc> is sent toward <nc>the browser application</nc> has <nc>a reduced number</nc> of <nc>browser links</nc> than <nc>the markup language content</nc> received from <nc>the content server</nc>; and wherein <nc>the modification</nc> is transparent to <nc>the browser application</nc> at <nc>the client computer</nc>.
| 5
|
5. <nc>The server</nc> of <nc>claim</nc> 1 , further comprising <nc>an object deliverer thread</nc> <nc>that</nc>: receives <nc>a browser link</nc> from <nc>the markup language object module</nc>; fetches <nc>the object</nc> associated with <nc>the browser link</nc>; and provide <nc>the fetched object</nc> to <nc>the markup language object module</nc>; and <nc>the markup language object module</nc> further to traverses <nc>the browser link</nc> by providing <nc>the browser link</nc> to <nc>the object deliverer thread</nc> and receiving <nc>the fetched object</nc> from <nc>the object deliverer thread</nc>.
|
8429171
|
12193896
|
1. <nc>A computer</nc> implemented <nc>information retrieval system</nc> for searching <nc>a corpus</nc>, <nc>the system</nc> comprising: <nc>a computer</nc>; <nc>a scoring module</nc> configured to, for <nc>each</nc> of <nc>one or more searches</nc> of <nc>at least a subset</nc> of <nc>the corpus</nc>, generate <nc>a confidence score</nc> for <nc>each</nc> of <nc>one or more putative occurrences</nc> of <nc>a search query</nc> in <nc>the at least a subset</nc> of <nc>the corpus</nc>; <nc>a threshold module</nc> configured to adjust <nc>a threshold</nc> to maintain <nc>a consistent user experience</nc> across <nc>the one or more searches</nc> according to <nc>a consistency criterion</nc>, wherein <nc>the threshold</nc> is adjusted according to at least one of <nc>a duration</nc> of <nc>the at least a subset</nc> of <nc>the corpus</nc> or <nc>an audio quality</nc> of <nc>the at least a subset</nc> of <nc>the corpus</nc>; and <nc>a display module</nc> configured to display <nc>putative occurrences</nc> of <nc>the search query</nc> having <nc>a confidence score</nc> greater than <nc>the threshold</nc>; wherein <nc>the consistency criterion</nc> is selected from <nc>the group</nc> consisting of <nc>a constant number</nc> of <nc>false alarms</nc> in <nc>the putative occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>, <nc>a constant false alarm rate</nc> in <nc>the putative occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>, <nc>a constant number</nc> of <nc>true occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>, <nc>a constant precision value</nc> in <nc>the putative occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>, <nc>a constant number</nc> of <nc>missed occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>, and <nc>a constant recall value</nc> in <nc>the putative occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>.
| 5
|
5. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the consistency criterion</nc> comprises <nc>a constant precision value</nc> in <nc>the putative occurrences</nc> of <nc>the search query</nc> for <nc>each</nc> of <nc>the one or more searches</nc>.
|
9799312
|
15179055
|
1. <nc>A method</nc> implemented by <nc>an information handling system</nc> <nc>that</nc> includes <nc>a memory</nc> and <nc>a processor</nc>, <nc>the method</nc> comprising: configuring <nc>a reinforcement learning model</nc> based on <nc>one or more inspiration selections</nc> received from <nc>a user</nc>, <nc>each</nc> of <nc>the one or more inspiration selections</nc> corresponding to <nc>one or more musical characteristics</nc>, wherein <nc>the</nc> configuring <nc>further comprises</nc>: determining <nc>one or more emotion characteristics</nc> of <nc>one or more songs</nc> corresponding to at least one of <nc>the one or more inspiration selections</nc>; displaying <nc>one or more emotion objects</nc> to <nc>the user</nc> on <nc>a display</nc>, <nc>each</nc> of <nc>the one or more emotion objects</nc> corresponding to one of <nc>the one or more emotion characteristics</nc>; receiving <nc>at least one emotion</nc> <nc>object adjustment</nc> from <nc>the user</nc> <nc>that</nc> adjusts <nc>a size</nc> of at least one of <nc>the one or more emotion objects</nc>; adjusting at least one of <nc>the one or more musical characteristics</nc> based on <nc>the emotion object adjustment</nc>; and loading <nc>the at least one adjusted musical characteristic</nc> into <nc>an environment</nc> in <nc>the reinforcement learning model</nc> to adjust <nc>a reward structure</nc> of <nc>the environment</nc>; performing <nc>a plurality</nc> of <nc>training iterations</nc> using <nc>the configured reinforcement learning model</nc>, wherein <nc>the plurality</nc> of <nc>training iterations</nc> generate <nc>a plurality</nc> of <nc>actions</nc> and <nc>the environment</nc> generates <nc>a plurality</nc> of <nc>rewards</nc> corresponding to <nc>the plurality</nc> of <nc>actions</nc>; and generating <nc>a musical composition</nc> based on <nc>the plurality</nc> of <nc>actions</nc> in <nc>response</nc> to determining that <nc>the plurality</nc> of <nc>rewards</nc> reach <nc>an empirical threshold</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 wherein at least one of <nc>the one or more inspiration selections</nc> is <nc>theme selection</nc>, <nc>the method</nc> further comprising: identifying <nc>the one or more musical characteristics</nc> of <nc>an ambience</nc> corresponding to <nc>the theme selection</nc>; and loading <nc>the one or more musical characteristics</nc> into <nc>the environment</nc> of <nc>the reinforcement learning model</nc>, wherein <nc>the environment</nc> generates <nc>the plurality</nc> of <nc>rewards</nc> based on <nc>the one or more musical characteristics</nc>.
|
8977553
|
13562228
|
1. <nc>A method</nc> comprising: receiving <nc>an indication</nc> of <nc>user activity</nc> with <nc>respect</nc> to <nc>a text input system</nc> of <nc>an electronic device</nc>; determining <nc>one or more activity indicators</nc> corresponding to <nc>user typing habits</nc> based on at least <nc>the user activity</nc>, including typing <nc>speed</nc>; identifying <nc>one or more components</nc> of <nc>the text input system</nc>, <nc>each component</nc> providing <nc>a typing assistance functionality</nc> to <nc>a user</nc> and being associated with <nc>a set</nc> of <nc>parameters</nc>, <nc>the one or more components</nc> including <nc>an auto-correct component</nc> and <nc>a predictive text component</nc>; and for <nc>each</nc> of <nc>the one or more components</nc>: determining, by <nc>a processor</nc>, whether <nc>the component</nc> should be adjusted based on <nc>the one or more activity indicators</nc>; and dynamically adjusting <nc>the component</nc> responsive to determining that <nc>the component</nc> should be adjusted based on <nc>the one or more activity indicators</nc>; wherein dynamically adjusting <nc>the auto-correct component</nc> and <nc>the predictive text component</nc> comprises adjusting <nc>a ratio</nc> of <nc>display</nc> of <nc>predictive text</nc> to <nc>auto-correct text</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the one or more components</nc> comprise <nc>a dictionary component</nc> for providing <nc>a dictionary</nc> in <nc>association</nc> with <nc>the text input system</nc>, and wherein <nc>the set</nc> of <nc>parameters</nc> associated with <nc>the dictionary component</nc> comprises at least one of <nc>a parameter</nc> for <nc>a language</nc> of <nc>the dictionary</nc> or <nc>a priority</nc> for <nc>each</nc> of <nc>multiple language dictionaries</nc>.
|
8180647
|
13230254
|
1. <nc>A method</nc> comprising: identifying <nc>a communicative goal</nc> in <nc>a speech input</nc> of <nc>a user</nc>; generating, via <nc>a processor</nc>, a plurality of <nc>sentence plans</nc> based on <nc>the communicative goal</nc> and <nc>a dialog history database</nc>, wherein <nc>each sentence plan</nc> of <nc>the plurality</nc> of <nc>sentence plans</nc> is <nc>a viable and potentially usable prompt</nc> in <nc>response</nc> to <nc>the speech input</nc> of <nc>the user</nc>; ranking <nc>the plurality</nc> of <nc>sentence</nc> plans to yield <nc>ranked sentence plans</nc>; and outputting <nc>at least one ranked sentence plan</nc> to yield <nc>an output sentence plan</nc>, wherein <nc>the output sentence plan</nc> provides <nc>a best sentence plan</nc> for generating <nc>a realization</nc> for <nc>the communicative goal</nc> from among <nc>the ranked sentence plans</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the communicative goal</nc> is generated using <nc>recognition</nc> and understanding <nc>data</nc> from <nc>an automated dialog system</nc>.
|
9589011
|
14804234
|
1. <nc>A method</nc> comprising, by <nc>a computing device</nc>: accessing <nc>a prior structured query</nc> previously selected by <nc>a first user</nc> of <nc>an online social network</nc>, <nc>the prior structured query</nc> corresponding to <nc>a first set</nc> of <nc>search results</nc>, wherein <nc>the prior structured query</nc> comprises <nc>references</nc> to <nc>one or more objects</nc> associated with <nc>the online social network</nc>; identifying <nc>changes</nc> to <nc>the first set</nc> of <nc>search results</nc> corresponding to <nc>the prior structured query</nc>; and sending, to <nc>a client system</nc> of <nc>the first user</nc>, <nc>one or more suggested structured queries</nc> for <nc>display</nc> to <nc>the first user</nc>, wherein at least one of <nc>the suggested structured queries</nc> is <nc>a dynamic query</nc> comprising <nc>at least a portion</nc> of <nc>the prior structured query</nc> and <nc>a reference</nc> to <nc>the identified changes</nc> to <nc>the first set</nc> of <nc>search results</nc> corresponding to <nc>the prior structured query</nc>.
| 17
|
17. <nc>The method</nc> of <nc>claim</nc> 1 , <nc>wherein the one or more changes</nc> to <nc>the first set</nc> of <nc>search results</nc> comprise <nc>changes</nc> performed by <nc>one or more second users</nc> within <nc>a threshold degree</nc> of <nc>separation</nc> of <nc>the first user</nc> on <nc>the online social network</nc>.
|
8156117
|
12676216
|
1. <nc>A system</nc> for storing, searching and <nc>retrieval</nc> of <nc>a plurality</nc> of <nc>information objects</nc> of <nc>an arbitrary application domain</nc>, comprising: <nc>a distributed computer system</nc> comprising one or <nc>a plurality</nc> of <nc>computing devices</nc> connected with each other by <nc>communication lines</nc>, <nc>a connected logical storage network</nc>, wherein <nc>each node</nc> is <nc>an active unit</nc> of <nc>storage</nc> (<nc>AUS</nc>) and <nc>connections</nc> between <nc>nodes</nc> of <nc>said network</nc> are formed by <nc>links</nc> of <nc>one active units</nc> of <nc>storage</nc> to <nc>others</nc>, wherein <nc>every active unit</nc> of <nc>storage</nc> resides on one of <nc>the computing devices</nc> of <nc>said distributed computer system</nc> and comprises: at least one of <nc>said plurality</nc> of <nc>information objects</nc> (<nc>IOs</nc>), <nc>each</nc> of <nc>which</nc> is represented in <nc>a tree-like structure</nc>, <nc>a list</nc> of <nc>links</nc> to <nc>a certain plurality</nc> of <nc>other active units</nc> of <nc>storage</nc> by <nc>means</nc> of <nc>which</nc> said <nc>AUS</nc> participates in <nc>the operation</nc> of <nc>the logical storage network</nc>, and <nc>an associated program agent</nc> <nc>that</nc> allows performing <nc>operations</nc> on said <nc>AUS</nc> in <nc>connection</nc> with searching, storing and retrieving <nc>information</nc> by <nc>user requests</nc> using <nc>said list</nc> of <nc>links</nc>, wherein <nc>a program agent</nc> of <nc>each active unit</nc> of <nc>storage</nc> compares <nc>the IO</nc> incorporated into <nc>it</nc> with <nc>the IO</nc> of <nc>any other AUS</nc> and based on <nc>the comparison results</nc> computes <nc>the value</nc> of <nc>metric distance</nc> between <nc>the compared IOs</nc> and <nc>the IOs</nc> are <nc>electronic documents</nc> in <nc>the form</nc> of <nc>XML documents</nc>.
| 27
|
27. <nc>A method</nc> for searching and retrieving <nc>information</nc> about <nc>objects</nc> of <nc>an arbitrary application domain</nc> in <nc>the system</nc> for <nc>storage</nc> and <nc>retrieval</nc> of <nc>a plurality</nc> of <nc>information objects</nc> according to <nc>claim</nc> 1 , comprising <nc>the steps</nc> of: a) generating <nc>a search mask</nc> comprising <nc>information</nc> relevant to <nc>the information</nc> to be retrieved from <nc>said system</nc>, and said <nc>search mask</nc> being <nc>an information object</nc> (<nc>IO</nc>) with <nc>a tree structure</nc> similar to <nc>the tree structure</nc> of <nc>information objects</nc> stored in <nc>said system</nc>, <nc>b</nc>) generating based on said <nc>search mask</nc> <nc>an active query</nc> object (AQO) comprising <nc>search criteria</nc>, c) comparing <nc>the AQO</nc> with <nc>IOs</nc> embedded in <nc>a each</nc> of <nc>the plurality</nc> of <nc>active units</nc> of <nc>storage</nc> on <nc>the logical storage network</nc> of <nc>said system</nc>, wherein moving from <nc>AUS</nc> to <nc>AUS</nc> following <nc>existing logical connections</nc> between <nc>them</nc> is performed in <nc>the direction</nc> of decreasing <nc>metric distance</nc> between <nc>the information objects</nc> incorporated into <nc>AUS</nc>, and d) retrieving <nc>an IO</nc> meeting <nc>the search criteria</nc> specified in <nc>the AQO</nc>.
|
8165899
|
11604460
|
1. <nc>A system</nc> for managing <nc>form-generated data</nc> related to <nc>a patient encounter</nc>, <nc>the system</nc> comprising: <nc>a form</nc> having designated <nc>information fields</nc> at <nc>different locations</nc> on <nc>the form</nc>; an electronic writing system configured to generate <nc>location information</nc> <nc>that</nc> identifies <nc>the location</nc> of <nc>a user</nc> writing on <nc>the form</nc>; <nc>a contextualizer</nc> configured to translate <nc>location information</nc> related to <nc>the user</nc> writing to <nc>a contextualized data element</nc>, wherein <nc>the contextualized data element</nc> comprises <nc>contextual information</nc> <nc>that</nc> is associated with <nc>the user writing</nc>, wherein <nc>the contextualizer</nc> includes <nc>a mapping data</nc> set that maps <nc>user areas</nc> on <nc>the form</nc> to <nc>labels</nc> <nc>that</nc> are associated with <nc>the designated information fields</nc> and wherein <nc>the contextualizer</nc> is configured to identify <nc>a label</nc> from <nc>the location information</nc> by comparing <nc>the location information</nc> to the mapping data set, <nc>the contextualized data element</nc> comprising <nc>the label</nc>; and an Electronic Medical Record (EMR)/Electronic Health Record (<nc>EHR</nc>) application <nc>that</nc> utilizes <nc>the label</nc> in <nc>the contextualized data element</nc> to perform <nc>a function</nc> <nc>that</nc> is related to <nc>the user</nc> writing on <nc>the form</nc>; wherein <nc>the contextualized data element</nc> is distributed to <nc>the EMR/EHR application</nc> via <nc>a publish/subscribe protocol</nc> in <nc>which</nc> <nc>the EMR/EHR application</nc> subscribes to <nc>a specific contextualized data element</nc> by identifying <nc>the label</nc> associated with <nc>the contextualized data element</nc>.
| 7
|
7. <nc>The system</nc> of <nc>claim</nc> 1 wherein <nc>the label</nc> comprises <nc>a standardized medical code</nc>.
|
9348925
|
13161836
|
1. <nc>A method</nc> performed by <nc>one or more data processing apparatuses</nc>, <nc>the method</nc> comprising: identifying <nc>a general search query</nc> <nc>that</nc> does not include <nc>a location phrase</nc>, wherein <nc>a location phrase</nc> is <nc>one or more terms</nc> <nc>that</nc> specify <nc>a geographic location</nc>; determining, for <nc>the general search query</nc>, <nc>a map query rate</nc> based on <nc>a ratio</nc> of <nc>a number</nc> of <nc>times</nc> that <nc>the general search query</nc> was received through <nc>an online map interface</nc> presenting <nc>the geographic location</nc> relative to <nc>a total number</nc> of <nc>times</nc> that <nc>the general search query</nc> was received; determining that <nc>the general search query</nc> is <nc>a locally significant query</nc> for <nc>the geographic location</nc> based, at least in <nc>part</nc>, on <nc>the map query rate</nc> for <nc>the general search query</nc> exceeding <nc>a threshold value</nc>; creating <nc>a local search query</nc> using <nc>the general search query</nc> and a location phrase representing <nc>the geographic location</nc>; requesting <nc>a set</nc> of <nc>general search results</nc> responsive to <nc>the general search query</nc> and <nc>a set</nc> of <nc>local search results</nc> responsive to <nc>the local search query</nc>; selecting, from <nc>the set</nc> of <nc>general search results</nc> and <nc>the set</nc> of <nc>local search results</nc>, <nc>a final set</nc> of <nc>search results</nc> including <nc>at least one local search result</nc> from <nc>the set</nc> of <nc>local search results</nc> and <nc>at least one general search result</nc> from <nc>the set</nc> of <nc>general search results</nc>; and providing <nc>data</nc> <nc>that</nc> cause <nc>presentation</nc> of <nc>the final set</nc> of <nc>search results</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein determining that <nc>the general search query</nc> is <nc>a locally significant query</nc> for <nc>the geographic location</nc> further comprises determining that <nc>the general search query</nc> has been received from <nc>user devices</nc> associated with <nc>the geographic location</nc> at least a threshold number of <nc>times</nc> over <nc>a specified period</nc>.
|
8392820
|
12628513
|
1. <nc>A method</nc> of establishing <nc>a plain text document</nc> from <nc>a HTML document</nc>, comprising <nc>the steps</nc> of<nc>: (A</nc>) acquiring <nc>a HTML document</nc> defined by <nc>HTML elements</nc>, <nc>each HTML element</nc> composed of <nc>tags</nc> and <nc>content</nc> between <nc>the tags</nc>; <nc>(B</nc>) pre-processing <nc>the HTML document</nc> by omitting <nc>some</nc> of <nc>the HTML elements</nc>, whereby <nc>the rest</nc> of <nc>the HTML document</nc> comprises <nc>at least one target tag</nc> and <nc>at least one corresponding content</nc>; <nc>(C</nc>) using <nc>a data structure</nc> to store <nc>the remaining tags</nc> of <nc>the pre-processed HTML document</nc>; <nc>(D</nc>) grouping <nc>the remaining HTML elements</nc> with <nc>the remaining tags</nc> stored in <nc>the data structure</nc> of <nc>the pre-processed HTML document</nc> into <nc>at least one target group</nc> according to <nc>the target</nc> tag(s), <nc>the step</nc> (<nc>D</nc>) further comprises <nc>the steps</nc> of: (D- 11 ) sequentially searching for <nc>a first content</nc> near <nc>the target tag</nc> from <nc>the rest</nc> of <nc>the HTML document</nc>, and identifying <nc>the first content</nc> as <nc>a first base content</nc>; (D- 12 ) sequentially searching for <nc>next content</nc> near <nc>the target tag</nc> from <nc>the first base content</nc>, and if there is <nc>no next content</nc> near <nc>the target tag</nc>, implementing <nc>the step</nc> (D- 15 ); (D- 13 ) if <nc>an interval</nc> between <nc>the next content</nc> of <nc>the step</nc> (D- 12 ) and <nc>the base content</nc> is smaller than <nc>a predetermined threshold</nc>, identifying <nc>the next content</nc> of <nc>the step</nc> (D- 12 ) as <nc>a current base content</nc>, and repeating <nc>the step</nc> (D- 12 ), otherwise, implementing <nc>the step</nc> (D- 14 ); (D- 14 ) grouping <nc>the first content</nc> and <nc>the current base</nc> <nc>content(s</nc>) into <nc>a target group</nc>, and identifying <nc>the next content</nc> as <nc>another first base content</nc>, implementing <nc>the step</nc> (D- 12 ); and (D- 15 ) grouping <nc>the first base content</nc> into one of <nc>the target groups</nc>; and <nc>(E</nc>) identifying <nc>the target</nc> <nc>group(s</nc>) most related to <nc>a title</nc> of <nc>the HTML document</nc> by comparing <nc>correlation(s</nc>) between <nc>the target</nc> <nc>group(s</nc>) and <nc>the title</nc>, and establishing <nc>a plain text document</nc> having <nc>the content</nc> of <nc>the identified target group</nc>.
| 13
|
13. <nc>The method</nc> of establishing <nc>a plain text document</nc> from <nc>a HTML document</nc> as claimed in <nc>claim</nc> 1 , wherein <nc>the target tags</nc> comprise <nc>tags</nc> <p> and <<nc>br</nc>>.
|
9002857
|
12461517
|
1. <nc>A method</nc> of <nc>scoring items</nc> resulting from <nc>a query</nc> containing <nc>specified terms</nc>, wherein <nc>the items</nc> being scored are annotated to <nc>a set</nc> of <nc>terms</nc> in <nc>one or more ontologies</nc> comprising <nc>the steps</nc> of: determining whether <nc>the set</nc> of <nc>terms</nc> in <nc>the one or more ontologies</nc> to <nc>which</nc> <nc>the items</nc> being scored are annotated are semantically related to <nc>the terms</nc> in <nc>the specified query</nc>, assigning <nc>an observed semantic similarity score</nc> to <nc>each</nc> of <nc>the items</nc> being scored; for <nc>each item</nc> being scored, determining <nc>the probability</nc> of obtaining <nc>the observed semantic similarity score</nc> in <nc>the event</nc> <nc>a random set</nc> of <nc>query terms</nc> is used instead of <nc>the specified query terms</nc>, assigning <nc>an individual P-value</nc> to <nc>the item</nc> being scored based on <nc>the determined probability</nc>, and scoring <nc>each</nc> of <nc>the items</nc>, using <nc>a computer</nc>, according to <nc>the assigned P-values</nc>.
| 15
|
15. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the P-values</nc> are compared to <nc>an adjustable P-value cutoff</nc>, and wherein <nc>items</nc> from <nc>the plurality</nc> of <nc>items</nc> are returned <nc>whose P-values</nc> pass <nc>the adjustable P-value cutoff</nc>.
|
8224849
|
13092083
|
1. <nc>A computer-readable storage device</nc> storing <nc>computer-executable instructions</nc> for controlling <nc>a computing device</nc> to identify <nc>images</nc> of <nc>a collection</nc> <nc>that</nc> are similar to <nc>a target image</nc>, by <nc>a method</nc> comprising: for <nc>each</nc> of <nc>a plurality</nc> of <nc>features</nc>, providing <nc>a cluster index data structure</nc> for <nc>the collection</nc> of <nc>images</nc>, <nc>the cluster index data structure</nc> defining <nc>clusters</nc> of <nc>images</nc> <nc>that</nc> are feature similar based on <nc>the values</nc> of <nc>that feature</nc>, such that for <nc>each feature</nc>, <nc>the images</nc> in <nc>the collection</nc> are clustered differently based on <nc>the values</nc> for <nc>that feature</nc>; for <nc>each</nc> of <nc>the plurality</nc> of <nc>features</nc>, identifying, from <nc>the cluster index data structure</nc> for <nc>that feature</nc>, <nc>candidate images</nc> <nc>that</nc> are feature similar to <nc>the target image</nc> based on <nc>that feature</nc>, <nc>the cluster index data structure</nc> defining, for <nc>each the plurality</nc> of <nc>features</nc> of <nc>images</nc>, <nc>clusters</nc> of <nc>images</nc> <nc>that</nc> are feature similar based on <nc>that feature</nc>; and for <nc>each</nc> of <nc>the candidate images</nc>, indicating <nc>similarity</nc> of <nc>that candidate image</nc> to <nc>the target image</nc> based on <nc>the features</nc> for <nc>which</nc> <nc>that candidate image</nc> is <nc>feature</nc> similar to <nc>the target image</nc>.
| 5
|
5. <nc>The computer-readable storage device</nc> of <nc>claim</nc> 1 wherein <nc>the candidate image</nc> is <nc>feature</nc> similar to <nc>the target object</nc> based on <nc>the values</nc> of <nc>the features</nc>.
|
9971762
|
15525800
|
1. <nc>A computer</nc> implemented <nc>method</nc> for detecting <nc>meaningless lexical units</nc> in <nc>an electronic message</nc> received by <nc>a server</nc> as to generate <nc>an abstract</nc> of <nc>the electronic message</nc>, <nc>the method</nc> comprising: <nc>(i</nc>) performing, by <nc>the server</nc>, <nc>a syntax analysis</nc> of <nc>a most significant part</nc> of <nc>the electronic message</nc> and determining <nc>at least one lexical unit</nc> as <nc>a first potential meaningless lexical unit</nc>, <nc>the first potential meaningless lexical unit</nc> comprising <nc>a plurality</nc> of <nc>control elements</nc>, <nc>the most significant part</nc> of <nc>the electronic message</nc> having been determined by analyzing <nc>a most significant logical block</nc> of <nc>source code</nc> from <nc>a plurality</nc> of <nc>the logical blocks</nc> of <nc>a source code</nc> of <nc>the electronic message</nc>; (ii) determining, by <nc>the server</nc>, <nc>a numerical control sum</nc> of <nc>the first potential meaningless lexical unit</nc>, <nc>the numerical control sum</nc> being based on <nc>the plurality</nc> of <nc>control elements</nc> having <nc>a respective numeric value representative</nc> of <nc>the first potential meaningless lexical unit</nc>; (iii) using <nc>the numerical control sum</nc>, accessing <nc>a lexical unit database</nc> located on <nc>the server</nc>, <nc>the lexical unit database</nc> containing <nc>a plurality</nc> of <nc>pre-determined meaningless lexical units</nc> with <nc>their associated pre-determined numerical control sums</nc>; (iv) determining, by <nc>the server</nc>, <nc>the first potential meaningless lexical unit</nc> is <nc>a meaningless lexical unit</nc> if <nc>the lexical units database</nc> includes <nc>at least one meaningless lexical unit</nc> with <nc>the pre-determined numerical control sum</nc> matching <nc>the numerical control sum</nc> of <nc>the first potential meaningless lexical unit</nc>, <nc>the matching comprising</nc> checking <nc>a measure</nc> of <nc>a difference</nc> between <nc>the numerical control sums</nc> and determining <nc>the numerical control sums</nc> as matching if <nc>the measure</nc> of <nc>the difference</nc> is within <nc>a predefined permissible amplitude</nc> of <nc>the difference</nc>; (v) if <nc>the first potential meaningless lexical unit</nc> is determined to be <nc>the meaningless lexical unit</nc>, generating, by <nc>the server</nc>, <nc>the abstract</nc> of <nc>the electronic message</nc>, <nc>the abstract</nc> not including <nc>the at least one meaningless lexical unit</nc>.
| 12
|
12. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the electronic message</nc> is <nc>an e-mail message</nc>.
|
9946749
|
15260136
|
1. A computer-implemented method comprising: generating, for <nc>an original attribute</nc> of <nc>a relation</nc> maintained in <nc>a database system</nc>, <nc>a respective auxiliary attribute</nc> for <nc>each interval size</nc> of <nc>a plurality</nc> of <nc>interval sizes</nc>, <nc>each interval size</nc> corresponding to <nc>a different respective power</nc> of <nc>a particular exponent base</nc>; computing, for <nc>each data entry</nc> of <nc>the relation</nc> and for <nc>each interval size</nc> of <nc>the plurality</nc> of <nc>interval sizes</nc>, a respective interval number for <nc>the interval size</nc> to <nc>which</nc> <nc>an original attribute value</nc> of <nc>the data entry</nc> belongs; storing <nc>each respective computed interval number</nc> for <nc>each data entry</nc> in <nc>the database system</nc> as <nc>an auxiliary attribute value</nc> of <nc>a corresponding auxiliary attribute</nc> for <nc>the data entry</nc>; receiving, by <nc>a query rewriter</nc> of <nc>a user device</nc> in <nc>communication</nc> with <nc>the database system</nc>, <nc>an original query</nc> having <nc>an inequality expression</nc> for <nc>the original attribute</nc>; generating <nc>a new query</nc> <nc>that</nc> replaces <nc>the inequality expression</nc> with <nc>multiple equality expressions</nc>, wherein <nc>each equality expression</nc> references <nc>a different respective auxiliary attribute</nc>, each auxiliary attribute representing <nc>a different respective interval size</nc> for <nc>values</nc> of <nc>the original attribute</nc>; providing, by <nc>the user device</nc> to <nc>the database system</nc>, <nc>the new query</nc> having <nc>the multiple equality expressions</nc> instead of <nc>the original query</nc>; and receiving, by <nc>the user device</nc> from <nc>the database system</nc>, <nc>query</nc> results <nc>that</nc> satisfy <nc>the inequality expression</nc> for <nc>the original attribute</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein generating <nc>the new query</nc> <nc>that</nc> replaces <nc>the inequality expression</nc> with <nc>a bounded number</nc> of <nc>equality expressions</nc> comprises: generating <nc>a disjunct</nc> of <nc>equality expressions</nc> including determining, for <nc>each interval size</nc> of <nc>a plurality</nc> of <nc>interval sizes</nc> other than <nc>a maximum interval size</nc>, whether <nc>respective equality expressions</nc> <nc>that</nc> test for <nc>an auxiliary attribute value</nc> belonging to <nc>a first interval</nc> at <nc>the interval size</nc>, <nc>a last interval</nc> at <nc>the interval size</nc>, or <nc>both</nc>, should be added to <nc>the disjunct</nc> of <nc>equality expressions</nc>.
|
4818131
|
06947828
|
1. <nc>A typewriter</nc> having <nc>a function</nc> of automatically indicating <nc>a candidate correct word</nc> for <nc>a mispelled word</nc>, comprising: <nc>a keyboard</nc> having <nc>a multiplicity</nc> of <nc>keys</nc>; <nc>an input data memory</nc> for storing <nc>input data</nc> constituting <nc>words</nc> entered through <nc>said keyboard</nc>; <nc>a printing mechanism</nc> operable for printing <nc>characters</nc> corresponding to said <nc>input data</nc> entered through <nc>said keyboard</nc>; <nc>a display device</nc> operable for displaying said <nc>characters</nc> in <nc>only one line</nc>; and a control device connected to said <nc>keyboard</nc>, said <nc>input data memory</nc>, said <nc>printing mechanism</nc> and said <nc>display device</nc>, for controlling said <nc>input data memory</nc>, said <nc>printing mechanism</nc> and said <nc>display device</nc>, to store, print and display said <nc>input data</nc> entered through <nc>said keyboard</nc>; said <nc>control device</nc> including <nc>a dictionary memory</nc> for storing <nc>data</nc> of <nc>a multiplicity</nc> of <nc>words</nc>, and spell-checking and correct-word indicating means for sequentially comparing <nc>each</nc> of <nc>the words</nc> of <nc>said input data</nc> with <nc>said multiplicity</nc> of <nc>words</nc> stored in <nc>said dictionary memory</nc>, to check said <nc>input data</nc> for <nc>any misspelled wrong words</nc>, and if <nc>a wrong word</nc> is found in <nc>said input data</nc>, searching for <nc>at least one candidate correct word</nc> to be substituted for said found <nc>wrong word</nc>, from among <nc>said words</nc> stored in <nc>said dictionary memory</nc>, displaying said <nc>wrong word</nc>, and upon locating said <nc>candidate correct word</nc> displaying said candidate correct word in <nc>relation</nc> to said found <nc>wrong word</nc> on <nc>said display device</nc>, without <nc>operator intervention</nc> upon <nc>location</nc> of <nc>a candidate correct word</nc>.
| 5
|
5. <nc>A typewriter</nc> according to <nc>claim</nc> 1, wherein said <nc>keyboard</nc> further has <nc>a next key</nc>, and said <nc>spell-checking and correct-word indicating means</nc> is operable for indicating on <nc>said display device</nc> <nc>a plurality</nc> of <nc>candidate correct words</nc> for <nc>each wrong word</nc> found, one after <nc>another</nc> in <nc>a predetermined order</nc>, upon <nc>successive operations</nc> of said next key.
|
9363560
|
14661488
|
1. <nc>A method</nc>, comprising: causing, by <nc>a computing device</nc>, display of <nc>a blending</nc> of <nc>user-selectable content</nc> for <nc>a content category</nc> in <nc>at least two levels</nc> of <nc>a hierarchy</nc> of <nc>levels</nc>, the blending of <nc>user-selectable content</nc> for <nc>the content category</nc> comprising at least two of <nc>linear content</nc>, <nc>non-linear content</nc> or <nc>managed content</nc>, the at least two of <nc>the linear content</nc>, <nc>the non-linear content</nc> or <nc>the managed content</nc> being presented together within <nc>the at least two levels</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the blending</nc> of <nc>user-selectable content</nc> for <nc>the content category</nc> comprises <nc>linear content</nc> and <nc>non-linear content</nc>.
|
7902969
|
11837393
|
1. <nc>A method</nc> for <nc>use</nc> in <nc>a system</nc> in <nc>a vehicle</nc>, comprising: sensing <nc>an ambient temperature</nc> of <nc>a compartment</nc> of <nc>the vehicle</nc>; determining <nc>a body temperature</nc> of <nc>a user</nc> in <nc>the compartment</nc>; and invoking <nc>a test</nc> for testing <nc>the user</nc> when <nc>the body temperature</nc> of <nc>the user</nc> changes more than <nc>a predetermined value</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 wherein <nc>the body temperature</nc> is determined from <nc>time</nc> to <nc>time</nc>.
|
10027611
|
15074788
|
1. <nc>A method</nc> for processing <nc>electronic messages</nc> (<nc>e</nc>-<nc>mails</nc>) is provided, <nc>the method</nc> comprising: receiving <nc>an electronic message</nc>, wherein <nc>the received electronic message</nc> includes <nc>a sender's address</nc>; comparing <nc>the sender's address</nc> associated with <nc>the electronic message</nc> to <nc>sender addresses</nc> found on <nc>a blacklist</nc> <nc>that</nc> contains <nc>a plurality</nc> of <nc>sender addresses</nc> and <nc>signatures</nc> of <nc>previously sent spam</nc>; executing <nc>instructions</nc> stored in <nc>memory</nc>, wherein <nc>the instructions</nc> are executed by <nc>a processor</nc> to: receive <nc>user input</nc> <nc>that</nc> classifies <nc>the received electronic message</nc> when <nc>the sender's address</nc> for <nc>the received electronic message</nc> is not found within <nc>the blacklist</nc>, and update <nc>the blacklist</nc> based on <nc>the received user input</nc>, wherein updating <nc>the blacklist</nc> comprises: summarizing <nc>content</nc> of <nc>the received electronic message</nc>, wherein summarizing <nc>the content</nc> of <nc>the received electronic message</nc> includes identifying that <nc>the content</nc> includes <nc>at least one word</nc> <nc>that</nc> has <nc>a plurality</nc> of <nc>possible canonical equivalents</nc>, selecting one of <nc>the equivalents</nc> for <nc>the at least one word</nc> based on <nc>likely association</nc> with <nc>the spam</nc>, assigning <nc>a probability</nc> of <nc>occurrence</nc> to <nc>the at least one word</nc> in <nc>the received electronic message</nc>, and <nc>the probability</nc> of <nc>occurrence</nc> corresponding to <nc>a probability</nc> that <nc>the at least one word</nc> occurs in <nc>a language</nc>, generating <nc>one or more signatures</nc> for <nc>the received electronic message</nc> based on <nc>the summarized content</nc>, and associating <nc>the generated signatures</nc> with <nc>the sender's address</nc>, and process <nc>the received electronic message</nc> based on <nc>the blacklist</nc> and based on <nc>the selected equivalent</nc> most likely being associated with <nc>the spam</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , wherein summarizing <nc>content</nc> of <nc>the received electronic message</nc> includes replacing <nc>one or more words</nc> found in <nc>the received electronic message</nc> with <nc>the selected canonical equivalent</nc>.
|
9135912
|
13622547
|
1. <nc>A computer-implemented method</nc> comprising: identifying, from among <nc>a set</nc> of <nc>query terms</nc>, <nc>a particular query term</nc> that <nc>(i</nc>) does not occur in <nc>a lexicon</nc> of <nc>terms</nc>, and (ii) has <nc>no designated, canonical phonetic representation</nc> in <nc>a pronunciation phonetic dictionary</nc>, wherein <nc>a canonical phonetic representation</nc> comprises <nc>a sequence</nc> of <nc>phonemes</nc>; generating <nc>a phonetic representation estimate</nc> for <nc>the particular query term</nc> that <nc>(i</nc>) does not occur in <nc>the lexicon</nc> of <nc>terms</nc>, and (ii) has <nc>no designated, canonical phonetic representation</nc> in <nc>the pronunciation phonetic dictionary</nc>; transmitting <nc>data</nc> identifying <nc>at least a portion</nc> of <nc>a term</nc> <nc>that</nc> does occur in <nc>the lexicon</nc> of <nc>terms</nc> and <nc>the particular query term</nc> to <nc>a spelling correction server</nc>; receiving, from <nc>the spelling correction server</nc>, <nc>data</nc> <nc>that</nc> specifies <nc>a spelling correction confidence score</nc>, wherein <nc>the spelling correction confidence score</nc> reflects <nc>a probability</nc> that <nc>the term</nc> <nc>that</nc> does occur in <nc>the lexicon</nc> of <nc>terms</nc> is <nc>a correct spelling</nc> of <nc>the particular query term</nc>; determining that <nc>the spelling correction confidence score</nc> satisfies <nc>a predetermined threshold</nc>; and in <nc>response</nc> to determining that <nc>the spelling correction confidence score</nc> satisfies <nc>a predetermined threshold</nc>, designating, by <nc>one or more computing devices</nc>, <nc>the phonetic representation estimate</nc> for <nc>the particular query term</nc> as <nc>a canonical phonetic representation</nc>, in <nc>the phonetic dictionary</nc>, of <nc>the term</nc> <nc>that</nc> does occur in <nc>the lexicon</nc> of <nc>terms</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , comprising: receiving <nc>a data representation</nc> of <nc>an utterance</nc> in <nc>which</nc> <nc>a user</nc> has spoken <nc>the term</nc> <nc>that</nc> does occur in <nc>the lexicon</nc> of <nc>terms</nc>; and using <nc>the phonetic representation estimate</nc> for <nc>the particular query term</nc>, in outputting <nc>the term</nc> <nc>that</nc> does occur in <nc>the lexicon</nc> of <nc>terms</nc> as <nc>part</nc> of <nc>a transcription</nc> of <nc>the utterance</nc>.
|
9128980
|
14611232
|
1. A method of accessing <nc>data</nc>, including: accessing <nc>a data model structure</nc>, the data model structure comprising: <nc>a root object query</nc> <nc>that</nc>, when executed, returns a set of <nc>time</nc> stamped events in <nc>a data store</nc> on <nc>a computing device</nc>, <nc>each event</nc> including <nc>a portion</nc> of <nc>unstructured data</nc>; <nc>a model schema</nc> that references <nc>fields</nc> <nc>that</nc> can be extracted, by <nc>an extraction rule</nc> or <nc>regular expression</nc>, from <nc>the unstructured data</nc> in <nc>the time</nc> <nc>stamped events</nc> without modifying <nc>the unstructured data</nc>; and <nc>one or more submodels</nc>; <nc>each</nc> of <nc>the submodels</nc> comprising: <nc>a child object</nc> <nc>that</nc> provides for <nc>narrower search criteria</nc> than <nc>the root object query</nc> such that, when <nc>the child</nc> object query is executed against <nc>the time stamped events</nc>, <nc>the child object query</nc> returns <nc>a subset</nc> of <nc>the set</nc> of <nc>time</nc> <nc>stamped events</nc> <nc>that</nc> is smaller than <nc>the set</nc>; <nc>a submodel schema</nc> <nc>that</nc> inherits <nc>one or more fields</nc> referenced in <nc>the model schema</nc>; and <nc>the submodel</nc> schema further references <nc>additional fields</nc> <nc>that</nc> can be extracted, by <nc>an extraction rule</nc> or <nc>regular expression</nc>, from <nc>the unstructured data</nc> in <nc>the time</nc> <nc>stamped events</nc> without modifying <nc>the unstructured data</nc>; receiving electronically <nc>a data request</nc> comprising <nc>reference</nc> to <nc>a submodel</nc> selected from <nc>the data model structure</nc> and <nc>a query</nc> to be performed against <nc>the subset</nc> referenced by <nc>the selected submodel</nc>; and identifying <nc>responsive events</nc>, including extracting <nc>values</nc> from <nc>at least some</nc> of <nc>the events</nc> in <nc>the subset</nc> at <nc>query time</nc> using <nc>the extraction rule</nc> or <nc>regular expression</nc> in <nc>the submodel schema</nc> without modifying <nc>the unstructured event</nc> and matching <nc>the extracted values</nc> to <nc>the query</nc>; returning <nc>at least some values</nc> from or derived from <nc>the fields</nc> in <nc>the responsive events</nc> referenced by <nc>the submodel schema</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , further including causing <nc>display</nc> of <nc>the values</nc> from or derived from <nc>the fields</nc> in <nc>the responsive events</nc> as <nc>values</nc> in <nc>rows</nc> in <nc>a pivot report template</nc>.
|
10055501
|
14935174
|
1. <nc>A method</nc> for processing <nc>an electronic query</nc>, comprising: receiving <nc>an electronic query</nc> from <nc>a client computer</nc> at <nc>a server computer</nc>, wherein <nc>the server computer</nc> is configured for: analyzing <nc>the query</nc> using <nc>a language modeling engine</nc> and <nc>a knowledge base</nc> to compute <nc>match scores</nc> and to classify <nc>the query</nc> into <nc>one or more predefined categories</nc> stored in <nc>the knowledge base</nc> based upon <nc>the match scores</nc>, wherein <nc>each</nc> of <nc>the predefined categories</nc> is associated with <nc>a suggested response</nc>; wherein <nc>the language modeling engine</nc> analyzes <nc>a natural language text</nc> of <nc>the query</nc> to generate <nc>concepts</nc> associated with <nc>the query</nc>, statistically compares <nc>the concepts</nc> with <nc>rules</nc> associated with <nc>the rule-oriented nodes</nc> and with <nc>concepts</nc> associated with <nc>the concept-oriented nodes</nc> stored in <nc>the knowledge base</nc>, and computes <nc>the match scores</nc> for <nc>one or more concept-oriented nodes</nc> representing one or more of <nc>the predefined categories</nc>; determining if <nc>the query</nc> meets <nc>any</nc> of <nc>one or more predetermined threshold levels</nc> for <nc>an automated response</nc>, based upon <nc>the match scores</nc>; transmitting <nc>a suggested response page</nc> to <nc>the client computer</nc>, if <nc>the query</nc> does meet <nc>any</nc> of <nc>the predetermined threshold levels</nc> for <nc>the automated response</nc>, wherein <nc>the suggested response page</nc> includes <nc>the suggested response</nc> associated with <nc>each</nc> of <nc>the predefined categories</nc> with <nc>an associated match</nc> score greater than or equal to a corresponding one of <nc>the predetermined threshold levels</nc>; otherwise routing <nc>the query</nc> to <nc>an agent</nc> for <nc>further analysis</nc>, if <nc>the query</nc> does not meet <nc>any</nc> of <nc>the predetermined threshold levels</nc> for <nc>the automated response</nc>, wherein <nc>the client computer</nc> is sent <nc>a confirmation page</nc> confirming that <nc>the query</nc> is being routed to <nc>the agent</nc> for <nc>further analysis</nc>, and <nc>the agent</nc> subsequently replies to <nc>the query</nc>; and wherein a language analysis server processes <nc>the agent's reply</nc> to <nc>the client computer</nc> to generate <nc>agent-based feedback</nc>, and <nc>the language analysis server</nc> updates <nc>the knowledge base</nc> based upon <nc>the agent-based feedback</nc>; wherein <nc>the language analysis server</nc> modifies <nc>concepts</nc>, adds <nc>new concepts</nc>, eliminates <nc>concepts</nc>, or modifies <nc>weights</nc> assigned to <nc>different concepts</nc> associated with <nc>concept-oriented nodes</nc> stored in <nc>the knowledge base</nc>, based upon <nc>the agent-based feedback</nc>; and wherein <nc>the query</nc> is considered resolved, if <nc>the client computer</nc> selects <nc>a suggested response</nc> or if <nc>the client computer</nc> does not select <nc>any response</nc>; receiving <nc>client-based feedback</nc>, in <nc>response</nc> to <nc>the query</nc> being resolved, for <nc>use</nc> in updating <nc>the knowledge base</nc>, wherein: if <nc>the client computer</nc> selects <nc>a suggested response</nc> corresponding to <nc>a high match score</nc>, then <nc>the client computer</nc> generates <nc>a positive client-based feedback</nc> for <nc>use</nc> in updating <nc>the knowledge base</nc>, and if <nc>the client computer</nc> selects <nc>a suggested response</nc> corresponding to <nc>a low match score</nc>, then <nc>the client computer</nc> generates <nc>a negative client-based feedback</nc> for <nc>use</nc> in updating <nc>the knowledge base</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the suggested response</nc> includes <nc>a message</nc> <nc>that</nc> recites “<nc>no response</nc> was found”.
|
9886954
|
15282860
|
1. <nc>A method</nc>, comprising: receiving, at <nc>a processor</nc>, <nc>one or more context aware processing parameters</nc> and <nc>an ambient audio stream</nc>, wherein <nc>the ambient audio stream</nc> is transmitted from <nc>one or more active acoustic filters</nc> to <nc>the processor</nc>, wherein <nc>the one or more active acoustic filters</nc> are configured to convert <nc>ambient sound</nc> received by <nc>the one or more active acoustic filters</nc> into <nc>the ambient audio stream</nc>; identifying using <nc>a machine learning model</nc> <nc>one or more sound characteristics</nc> associated with <nc>the ambient audio stream</nc>, wherein <nc>the machine learning model</nc> is comprised of <nc>a first hierarchical level</nc> <nc>that</nc> is comprised of <nc>a plurality</nc> of <nc>first hierarchical level machine learning models</nc> configured to each identify a corresponding sound characteristic within <nc>the ambient audio stream</nc>, <nc>a second hierarchical level</nc> comprised of <nc>a second hierarchical level machine learning model</nc> configured to determine <nc>one or more real characteristics</nc> of <nc>the ambient audio stream</nc> based on <nc>an output</nc> of <nc>the plurality</nc> of <nc>first hierarchical level machine learning models</nc>, and <nc>a third hierarchical level</nc> <nc>that</nc> is comprised of <nc>a third hierarchical level machine learning model</nc>; determining using <nc>the third hierarchical level machine learning model</nc> of <nc>the machine learning model</nc> <nc>one or more actions</nc> to perform based on <nc>the one or more context aware processing parameters</nc>, <nc>the identified one or more real characteristics</nc> determined by <nc>the second hierarchical level machine learning model</nc>, and <nc>a corresponding recommendation metric</nc> associated with <nc>the one or more actions</nc>; and performing <nc>the one or more actions</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein identifying using <nc>a machine leaning model</nc> <nc>one or more sound characteristics</nc> within <nc>the ambient audio stream</nc> includes providing <nc>the ambient audio stream</nc> to <nc>the first hierarchical level</nc> of <nc>the machine learning model</nc>.
|
9323456
|
13992372
|
1. <nc>A multi-character continuous handwriting input method</nc>, comprising <nc>the following steps</nc>: <nc>Step</nc> 110 : touching <nc>a touch screen</nc> with <nc>a handwriting pen</nc> or <nc>a finger</nc> to start inputting <nc>a stroke</nc>; <nc>Step</nc> 120 : moving <nc>the handwriting pen</nc> or <nc>the finger</nc> on <nc>the touch screen</nc>; recording <nc>a stroke track</nc> and displaying <nc>the stroke track</nc> on <nc>a handwriting area</nc> on <nc>the touch screen</nc>; <nc>Step</nc> 130 : moving <nc>the handwriting pen</nc> or <nc>the finger</nc> away from <nc>the touch screen</nc>, <nc>current stroke input</nc> being finished; <nc>Step</nc> 140 : determining whether <nc>the currently written stroke</nc> and <nc>a previously input stroke</nc> belong to <nc>the same character</nc>; if yes, going to Step 150 ; otherwise, going to <nc>Step</nc> 170 ; wherein <nc>the determining</nc> whether <nc>the currently written stroke</nc> and <nc>a previously input stroke</nc> belong to <nc>the same character</nc> is performed according to <nc>a relationship</nc> between <nc>geometric position information</nc> of <nc>the currently written stroke and geometric position information</nc> of <nc>a character</nc> formed of <nc>all the previously input strokes</nc>, comprising <nc>the following steps</nc>: <nc>Step</nc> 141 : determining whether <nc>the current stroke</nc> is <nc>the first stroke</nc> inputted by <nc>the user</nc>; if yes, going to <nc>Step</nc> 146 ; otherwise, going to <nc>Step</nc> 142 ; Step 142 : determining whether <nc>the current stroke</nc> is <nc>a new character stroke</nc> on <nc>the right side</nc> of <nc>a previous stroke</nc>; if yes, going to Step 145 ; otherwise, going to Step 143 ; <nc>Step</nc> 143 : determining whether <nc>the current stroke</nc> overlaps <nc>another previously written stroke</nc>; if yes, going to Step 144 ; otherwise, going to Step 145 ; <nc>Step</nc> 144 : determining whether <nc>an overlapping degree</nc> between <nc>the current stroke</nc> and <nc>the previously written stroke</nc> is greater than <nc>a given threshold</nc>; if yes, going to <nc>Step</nc> 146 ; otherwise, going to Step 145 ; <nc>Step</nc> 145 : returning <nc>a determination result</nc> that <nc>the currently input stroke</nc> and <nc>the previously input stroke</nc> probably belong to <nc>the same character</nc>; and Step 146 : returning <nc>a determination result</nc> that <nc>the currently input stroke</nc> and <nc>the previously input stroke</nc> do not belong to <nc>the same character</nc>; <nc>Step</nc> 150 : determining whether <nc>a new stroke</nc> is inputted; if yes, going to Step 120 ; otherwise, going to <nc>Step</nc> 160 ; <nc>Step</nc> 160 : submitting <nc>a currently written character track</nc> to <nc>a recognition engine</nc> for <nc>recognition</nc>, and outputting <nc>a recognition result</nc>; going to <nc>Step</nc> 220 ; <nc>Step</nc> 170 : determining whether <nc>some character</nc> on <nc>the touch screen</nc> is dimmed; if yes, going to Step 180 ; otherwise, going to Step 190 , wherein <nc>the dimmed character</nc> refers to <nc>a previous handwritten character</nc> <nc>that</nc> has been written and recognized, and <nc>a stroke color</nc> thereof has been processed in <nc>Step</nc> 200 ; Step 180 : clearing <nc>a previous dimmed character</nc>; <nc>Step</nc> 190 : combining <nc>all strokes</nc> except <nc>the current stroke</nc> into <nc>a handwritten character</nc>, submitting <nc>the handwritten character</nc> to <nc>the recognition engine</nc> for <nc>recognition</nc>, and outputting <nc>a recognition result</nc>; Step 200 : dimming <nc>the stroke color</nc> of <nc>the handwritten character</nc> formed of <nc>all the strokes</nc> except <nc>the current stroke</nc>, or making <nc>colors</nc> of <nc>an (i+1) th character</nc> and <nc>an i th character</nc> different, wherein <nc>the character</nc> is defined as <nc>a dimmed character</nc>; and Step 210 : determining whether <nc>a new stroke</nc> is inputted; if yes, going to Step 120 ; otherwise, going to <nc>Step</nc> 160 ; and <nc>Step</nc> 220 : ending.
| 5
|
5. <nc>The multi-character continuous handwriting input method</nc> as in <nc>claim</nc> 1 , wherein <nc>the multi-character continuous handwriting input method</nc> is a left-to-<nc>right</nc> handwriting input method <nc>that</nc> inputs <nc>characters</nc> from <nc>the left-hand side</nc> to <nc>the right-hand side</nc>.
|
9652478
|
14291471
|
1. <nc>A method</nc> for generating <nc>a schema</nc> for <nc>data asset information</nc>, <nc>the method</nc> comprising: accessing <nc>complex type information</nc> corresponding to <nc>a logical relational data model</nc> <nc>that</nc> defines <nc>an organization</nc> of <nc>the data asset information</nc>, <nc>the logical relational data model</nc> including <nc>a parent entity</nc> and <nc>child entities</nc> corresponding to <nc>the parent entity</nc>; treating <nc>the complex type information</nc> to produce <nc>scrubbed complex type information</nc>, said treating of <nc>the complex type information</nc> including removing <nc>at least one foreign key</nc> from at least one of <nc>the child entities</nc>; translating <nc>the scrubbed complex type information</nc> to produce <nc>a hierarchical data model</nc> corresponding to <nc>the logical relational data model</nc>, <nc>the hierarchical data model</nc> including <nc>a plurality</nc> of <nc>containers</nc> respectively corresponding to <nc>the child entities</nc> of <nc>the logical relational data model</nc>, <nc>the treating</nc> and translating being carried out such that <nc>the at least one foreign key</nc> removed from <nc>the at least one child entity</nc> is omitted from <nc>a first level container</nc> in <nc>the hierarchical data model</nc> and is present in <nc>a second level container</nc> in <nc>the hierarchical data model</nc>, <nc>the first level container</nc> paralleling <nc>the child entity</nc> from <nc>which</nc> <nc>the foreign key</nc> is removed, <nc>the second level container</nc> being at <nc>a higher level</nc> in <nc>the hierarchical data model</nc> than <nc>that</nc> of <nc>the first level container</nc>; and generating <nc>a schema</nc> for <nc>the data asset information</nc> based upon <nc>the hierarchical data model</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein said treating of <nc>the complex type information</nc> comprises replacing <nc>data types</nc> in <nc>the complex type information</nc> corresponding to <nc>the logical relational data model</nc>.
|
9812129
|
15327057
|
1. <nc>A method</nc> for operating <nc>a motor vehicle operating device</nc> based on <nc>voice control</nc> with <nc>first and second operating steps</nc> <nc>that</nc> are based on each other in <nc>a device</nc> of <nc>a motor vehicle</nc>, comprising: <nc>a first vocabulary</nc>, provided for <nc>the first operating step</nc>, is set with <nc>a speech recognition device</nc> and on <nc>the basis</nc> of <nc>the set first vocabulary</nc>, <nc>a first recognition result</nc> is generated for <nc>a first speech input</nc> of <nc>a user</nc>, <nc>the first operating step</nc> is carried out based on <nc>the first recognition result</nc> and instead of <nc>the first vocabulary</nc>, <nc>a second vocabulary</nc>, <nc>which</nc> is at least partially different from <nc>the first vocabulary</nc>, is set for <nc>the speech recognition device</nc> and <nc>a second speech input</nc> of <nc>the user</nc> is detected, wherein <nc>a repetition device</nc> recognizes <nc>a correction request</nc> of <nc>a user</nc> with or after <nc>the detection</nc> of <nc>the second speech input</nc> and reactivates <nc>the first vocabulary</nc> for <nc>the speech recognition device</nc> instead of <nc>the second vocabulary</nc>, and <nc>a second recognition result</nc> is generated on <nc>the basis</nc> of <nc>the reactivated first vocabulary</nc> based on <nc>at least one part</nc> of <nc>the second speech input</nc>, and <nc>the first operating step</nc> is carried out one more time based on <nc>the second recognition result</nc>.
| 4
|
4. <nc>The method</nc> according to <nc>claim</nc> 1 , <nc>wherein the the first operating step</nc> and <nc>the reactivation</nc> of <nc>the first vocabulary</nc> is carried out by <nc>the operating device</nc> without <nc>the participation</nc> of <nc>the user</nc>.
|
8386998
|
12061723
|
1. A software development apparatus for developing <nc>application software</nc> based on <nc>an object model</nc> <nc>that</nc> requires <nc>security</nc> in <nc>a web service application</nc>, <nc>the apparatus</nc> comprising: <nc>a display unit</nc> <nc>that</nc> displays, in <nc>a class diagram</nc> of <nc>said application software</nc>, <nc>a security annotation</nc> for adding <nc>security requirements</nc> for <nc>a service</nc>; <nc>input</nc> means for inputting <nc>an input</nc> with <nc>said security annotation</nc>; transforming <nc>means</nc> for transforming <nc>said class diagram</nc> into <nc>a configuration model</nc> based on <nc>a markup language</nc>, said <nc>class diagram</nc> being described in <nc>a UML</nc>, and said <nc>security requirements</nc> being described using <nc>an extension</nc> of <nc>the UML</nc>; and <nc>configuration-file creating</nc> <nc>means</nc> for creating <nc>a configuration file</nc> based on <nc>said markup language</nc> by serializing said <nc>configuration model</nc> based on <nc>said markup language</nc>.
| 5
|
5. <nc>The software development apparatus</nc> according to <nc>claim</nc> 1 , said <nc>markup language</nc> being <nc>XML</nc>.
|
8676564
|
13591292
|
1. <nc>A computer system</nc> configured to find <nc>correspondence</nc> between <nc>terms</nc> in <nc>two different languages</nc>, the system comprising: <nc>a unit</nc> configured to create <nc>a technical term</nc> set in <nc>a first language</nc>, <nc>a general term</nc> set in <nc>the first language</nc>, <nc>a technical term</nc> set in <nc>a second language</nc> and <nc>a general term</nc> set in <nc>the second language</nc>; <nc>a storage unit</nc> for storing <nc>the term</nc> set of <nc>the first language</nc>, <nc>the general term</nc> set in <nc>the first language</nc>, the technical terms set in <nc>the second language</nc> and the general term in <nc>the second language</nc>; <nc>a unit</nc> configured to create <nc>at least two bipartite graphs</nc>, wherein <nc>the first bipartite graph</nc> connects <nc>the technical term</nc> set and the general term set of <nc>the first language</nc> to each other with <nc>links</nc> on <nc>the basis</nc> of <nc>corpus information</nc>, wherein <nc>the second bipartite graph</nc> connects <nc>the general term set and technical term</nc> set of <nc>the second language</nc> to each other with <nc>links</nc> on <nc>the basis</nc> of <nc>corpus information</nc>, and wherein <nc>each</nc> of <nc>the links</nc> is weighted by <nc>a degree</nc> of <nc>association</nc> between <nc>terms</nc>; <nc>a unit</nc> configured to create <nc>a third bipartite graph</nc> by creating <nc>links</nc> between <nc>general terms</nc> in <nc>the first language</nc> and <nc>general terms</nc> in <nc>the second language</nc> by using <nc>a translation dictionary</nc> between <nc>general terms</nc> in <nc>the first language</nc> and <nc>general terms</nc> in <nc>the second language</nc>, each of <nc>the links</nc> being weighted by <nc>a degree</nc> of <nc>association</nc> between <nc>terms</nc>; <nc>a unit</nc> configured to create <nc>an association matrix M</nc> including <nc>the bipartite graphs</nc> between <nc>the technical term</nc> set and <nc>the general term</nc> set connected in <nc>each respective language</nc> and <nc>the bipartite graph</nc> between <nc>the general terms</nc> in <nc>the first language</nc> and <nc>the general terms</nc> in <nc>the second language</nc>, wherein <nc>the association matrix M</nc> is normalized such that <nc>a sum</nc> of <nc>each row</nc> is equal to one; <nc>a unit</nc> configured to calculate <nc>a similarity matrix Q</nc> by <nc>calculation</nc> of <nc>an inverse matrix</nc> (I−cM) of <nc>a matrix</nc> in <nc>which</nc> <nc>only a portion</nc> corresponding to <nc>both</nc> of <nc>the general term</nc> set in <nc>the first language</nc> and <nc>the technical term</nc> set in <nc>the first language</nc> and <nc>a portion</nc> corresponding to <nc>both</nc> of <nc>the general term</nc> set in <nc>the second language</nc> and <nc>the technical term</nc> set in <nc>the second language</nc> are left, where <nc>c</nc> is <nc>a positive number</nc> smaller than one; and <nc>a unit</nc> configured to output <nc>correspondence</nc> between <nc>the technical term</nc> set in <nc>the first language</nc> and <nc>the technical term</nc> set in <nc>the second language</nc> on <nc>the basis</nc> of <nc>predetermined components</nc> of <nc>the similarity matrix Q</nc>, wherein said <nc>unit</nc> configured to create <nc>the association matrix M</nc> includes creating <nc>a set</nc> of <nc>submatrices</nc> comprising: <nc>a submatrix</nc> created on <nc>the basis</nc> of <nc>links</nc> between <nc>the technical term</nc> set and <nc>the general term</nc> set in <nc>a first language</nc> and <nc>its transposed submatrix</nc>, <nc>a submatrix</nc> created on <nc>the basis</nc> of <nc>links</nc> between <nc>the technical term</nc> set and <nc>the general term</nc> set in <nc>a second language</nc> and <nc>its transposed submatrix</nc>, and <nc>a submatrix</nc> created on <nc>the basis</nc> of <nc>links</nc> between <nc>the general term</nc> set in <nc>the first language</nc> and <nc>the general term</nc> set in <nc>the second language</nc> and <nc>its transposed submatrix</nc>.
| 6
|
6. <nc>The system</nc> according to <nc>claim</nc> 1 , wherein <nc>the technical term sets</nc> of <nc>the first and second languages</nc> comprise <nc>terms</nc> in <nc>an automobile field</nc>.
|
9740752
|
15173009
|
1. <nc>A computer-implemented method</nc> comprising: extracting, by <nc>a communication network</nc>, <nc>linguistic data</nc> from <nc>at least one type</nc> of <nc>communication</nc> between <nc>a user</nc> of <nc>the communication network</nc> and <nc>one or more additional users</nc> of <nc>the communication network</nc>; retrieving <nc>at least one characteristic</nc> of <nc>the user</nc> from <nc>a user profile</nc> of <nc>the user</nc> at <nc>the communication network</nc>; applying <nc>at least one statistical model</nc> to <nc>the extracted linguistic data</nc> and the at least one retrieved characteristics of <nc>the user</nc>, <nc>the at least one statistical model</nc> being determined by: determining <nc>one or more personality characteristics</nc> of <nc>a training set</nc> of <nc>users</nc>, <nc>the one or more personality characteristics</nc> being determined based on <nc>responses</nc> to <nc>one or more surveys</nc> received from <nc>the training set</nc> of <nc>users</nc>; and generating <nc>the at least one statistical model</nc> based on <nc>the determined one or more personality characteristics</nc> and <nc>linguistic data</nc> retrieved from <nc>user profiles</nc> associated with <nc>the training set</nc> of <nc>users</nc> at <nc>the communication network</nc>; selecting <nc>at least one personality characteristics</nc> for <nc>the user</nc>, the selected at least one personality characteristic being associated with <nc>at least a threshold value</nc> from <nc>the at least one statistical model</nc>; storing <nc>the at least one selected personality</nc> characteristic in <nc>the user profile</nc> of <nc>the user</nc>; and presenting <nc>content</nc> to <nc>the user</nc> based at least in <nc>part</nc> on <nc>the at least one selected personality characteristic</nc>.
| 9
|
9. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , wherein presenting <nc>content</nc> to <nc>the user</nc> based at least in <nc>part</nc> on <nc>the at least one selected personality characteristic comprises</nc>: selecting <nc>one or more recommendations</nc> for <nc>actions</nc> to <nc>the user</nc> based at least in <nc>part</nc> on <nc>the at least one selected personality characteristic</nc>; and presenting <nc>the selected one or more recommendations</nc> for <nc>actions</nc> to <nc>the user</nc>.
|
9063926
|
13595241
|
1. <nc>A method</nc> for determining <nc>variants</nc> of <nc>a text entity</nc>, comprising: parsing <nc>the text entity</nc> into <nc>semantic components</nc>; generating <nc>variants</nc> based on <nc>input</nc> from <nc>auxiliary information</nc> and <nc>user configuration information</nc> for <nc>each</nc> of <nc>the semantic components</nc>, wherein based on <nc>the type</nc> of <nc>entity</nc>, <nc>different semantic components</nc> for <nc>each entity type</nc> are determined based on <nc>domain ontology</nc> and <nc>the user configuration information</nc>; and recomposing <nc>the entity</nc> in <nc>different morphological forms</nc> from <nc>the different variants</nc> of <nc>the semantic components</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising <nc>parsing</nc> based on <nc>the type</nc> of <nc>the entity</nc>.
|
8166386
|
12044973
|
1. <nc>A method</nc> for producing <nc>a patent specification</nc>, comprising <nc>the steps</nc> of: inputting <nc>a title</nc>; entering <nc>a set</nc> of <nc>selection items</nc> for selecting <nc>the different files</nc> of <nc>invention</nc> from <nc>one type</nc> of <nc>electronic circuit</nc>, <nc>structural device</nc>, <nc>software method</nc> or <nc>biological chemistry</nc>; providing <nc>a computer-enabled graphic interface</nc> according to <nc>the selected type</nc>, inputting <nc>names</nc> and <nc>functions</nc> thereof using <nc>the graphic interface</nc>, wherein <nc>the graphic interface</nc> uses <nc>blocks</nc> to show <nc>connecting relationship</nc> among <nc>the inputted contents</nc>, and <nc>each block</nc> represents <nc>the user-input name</nc> and <nc>function</nc> or <nc>description</nc> thereof, so as to produce <nc>an output data section</nc>, <nc>the output data description</nc> changes responsive to <nc>the change</nc> of <nc>the connecting relationship</nc> of <nc>the blocks</nc>; allowing to input <nc>basic elements</nc> or <nc>units</nc> using <nc>the graphic interfaces</nc> based on <nc>the technical characteristics</nc> and forming <nc>the output data section</nc>, comprising: (a) combining <nc>predetermined texts</nc> and <nc>symbols</nc> of <nc>a basic element</nc> within <nc>the graphic interface</nc> having <nc>the name</nc> for forming <nc>a data unit</nc>; (b) combining <nc>the predetermined texts</nc> and <nc>symbols</nc> of <nc>another basic element</nc> within <nc>the graphic interface</nc> having <nc>the name</nc> for forming <nc>another data unit</nc>; (c) determining whether <nc>the descriptions</nc> of <nc>the names</nc> and <nc>functions</nc> within <nc>the graphic interfaces</nc> input by <nc>the user</nc> form <nc>the data unit</nc>; (d) combining <nc>the data units</nc> when <nc>the input names</nc> and <nc>functions</nc> form <nc>the data unit</nc>; (e) forming <nc>the output data section</nc> to be <nc>an independent claim</nc>; forming <nc>the output data section</nc> as <nc>a set</nc> of <nc>claims</nc>; <nc>operating determinations</nc> over <nc>the set</nc> of <nc>claims</nc>, comprising: (f) determining whether <nc>a negative description</nc> is presented in <nc>the data section</nc>, and deleting <nc>the negative description</nc> if <nc>the determination</nc> is positive<nc>; (g</nc>) determining whether <nc>a defining description</nc> indicating the minimum, maximum or comprising <nc>0%</nc>, <nc>100%</nc> is presented in <nc>the data section</nc>, further comprising <nc>a step</nc> of determining <nc>the description</nc> indicating <nc>the maximum</nc> or minimum whether <nc>it</nc> is understood by <nc>those</nc> <nc>who</nc> refer to <nc>this technical field</nc>, deleting <nc>the description</nc> if <nc>the determination</nc> is negative, and displaying <nc>an indicating frame</nc> for generating <nc>a warning</nc> to <nc>the user</nc>; (h) determining whether <nc>an indefinite description</nc> is presented in <nc>the data section</nc>, and deleting <nc>such description</nc> if <nc>the determination</nc> is positive; (i) determining whether <nc>a description</nc> contains <nc>relative standard or indefinite level</nc> is presented in <nc>the data section</nc>, and deleting <nc>such description</nc> if <nc>the determination</nc> is positive; (j) determining whether <nc>the data section</nc> satisfies <nc>the principle</nc> of <nc>“single sentence</nc>”, and if <nc>the determination</nc> is negative, modifying <nc>the whole description</nc> of <nc>the data section</nc> to correct <nc>a description</nc> in <nc>compliance</nc> with <nc>requirement</nc> of <nc>single sentence</nc>; (k) completing <nc>the output data section</nc>; inputting <nc>data</nc> into <nc>multiple sets</nc> of <nc>text areas</nc>, comprising: (l) inputting <nc>motivation</nc>, <nc>objectives</nc>, and <nc>solutions</nc>; (m) combining, transferring and arranging <nc>the data section</nc> of <nc>multiple sets</nc> of <nc>output data sections</nc> as <nc>the contents</nc> of <nc>invention</nc>; (n) inputting <nc>prior art</nc> and <nc>drawbacks</nc>, wherein <nc>the section</nc> of <nc>prior art</nc> has <nc>a text area</nc> to be filled with <nc>reference documents</nc>, <nc>a patent number</nc> to link to <nc>a patent search website</nc> via <nc>Internet</nc> and <nc>the text file</nc> downloaded from <nc>the patent search website</nc>, and <nc>the descriptive texts</nc> of <nc>the reference</nc> are extracted as <nc>the contents</nc> of <nc>the data section</nc> of <nc>prior art</nc>; (o) combining, transferring and arranging <nc>the data section</nc> of <nc>multiple sets</nc> of <nc>output data sections</nc> as <nc>the section</nc> of <nc>prior art</nc>; (p) inputting <nc>comparison</nc><nc>; (q) combining</nc>, transferring and arranging <nc>the data section</nc> of <nc>multiple sets</nc> of <nc>output data sections</nc> as <nc>detailed description</nc> of <nc>the invention</nc>; (r) arranging <nc>the text area</nc> of <nc>comparison</nc> to <nc>the last paragraph</nc> of <nc>the data section</nc> in <nc>the detailed description</nc> of <nc>the invention</nc>; collocating <nc>the input data</nc> into <nc>the multiple sets</nc> of <nc>text areas</nc> with <nc>the output data section</nc>, transferring and arranging <nc>the input description</nc>, thereby forming <nc>multiple sets</nc> of <nc>output data sections</nc>; and outputting <nc>a document</nc> having <nc>the multiple sets</nc> of <nc>output data sections</nc> as <nc>a patent specification</nc>.
| 3
|
3. <nc>The method</nc> for producing <nc>a patent specification</nc> according to <nc>claim</nc> 1 , wherein forming <nc>the claims description and patent specification</nc> is applied in <nc>an internet website</nc> <nc>which</nc> allows <nc>the user</nc> to select <nc>an item</nc> of <nc>new filling</nc> or <nc>an new application icon</nc>, <nc>the set</nc> of <nc>selection items</nc> for selecting the different filed of <nc>invention</nc> are showed up.
|
7788652
|
10624705
|
1. <nc>A method</nc>, implemented at least in <nc>part</nc> by <nc>a computing device</nc> comprising <nc>a processing unit</nc> and <nc>memory</nc>, of <nc>representing type information</nc> for <nc>a typed intermediate language</nc> via <nc>objects</nc> of <nc>classes</nc> in <nc>a class hierarchy</nc>, wherein <nc>the class hierarchy</nc> comprises <nc>at least one class</nc> and <nc>a plurality</nc> of <nc>sub</nc><nc>-</nc><nc>classes</nc> for representing <nc>different type classifications</nc>, <nc>the method</nc> comprising: with <nc>the computing device</nc>: instantiating <nc>one or more objects</nc> of one or more of <nc>the sub</nc><nc>-</nc><nc>classes</nc> of <nc>the hierarchy</nc>, wherein <nc>the one or more sub</nc><nc>-</nc><nc>classes</nc> represent <nc>classifications</nc> of <nc>types</nc> for <nc>the typed intermediate language</nc>; and storing <nc>information</nc> in <nc>the one or more objects</nc>; wherein <nc>the typed intermediate language</nc> is capable of representing <nc>a plurality</nc> of <nc>different programming languages</nc>, and wherein <nc>the one or more objects</nc> represent <nc>type information</nc> for <nc>instructions</nc> in <nc>the typed intermediate language</nc>; wherein <nc>the classifications</nc> of <nc>types</nc> comprises <nc>a primitive type</nc> associated with <nc>a primitive type size</nc>, and wherein <nc>the primitive type size</nc> is settable to represent <nc>a constant size</nc>, <nc>the primitive type size</nc> is settable to represent <nc>a symbolic size</nc>, and <nc>the primitive type size</nc> is settable to represent <nc>an unknown size</nc>; and wherein one of <nc>the sub</nc><nc>-</nc><nc>classes</nc> representing <nc>a primitive type</nc> represents <nc>an unknown type</nc>, wherein <nc>the unknown type</nc> can represent <nc>any type</nc>, wherein <nc>the unknown type</nc> represents <nc>a lack</nc> of <nc>all type information</nc>, wherein <nc>a compiler</nc> drops <nc>type information</nc> by changing <nc>a known type</nc> to <nc>the unknown type</nc> during <nc>a stage</nc> of <nc>lowering</nc>, wherein <nc>the unknown type</nc> is set independently of <nc>the primitive type size</nc>, and wherein <nc>the classifications</nc> of <nc>types</nc> support <nc>an unknown type</nc> with <nc>an unknown primitive type size</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 wherein at least one of <nc>the one or more sub</nc><nc>-</nc><nc>classes</nc> inherits from <nc>an abstract type</nc> <nc>that</nc> wraps <nc>an externally defined type</nc>, the abstract type providing <nc>a mapping</nc> from <nc>the typed intermediate language</nc> to <nc>original source code</nc>.
|
7685088
|
11447546
|
1. <nc>A computing system</nc> for creating <nc>new concepts</nc> in <nc>existing ontologies</nc> based on <nc>new concept descriptions</nc> expressed in <nc>a natural language format</nc>, wherein <nc>words</nc> in <nc>a natural language</nc> are transformed into <nc>a certain language</nc> for expressing <nc>the new concepts</nc> in <nc>the existing ontologies</nc>, <nc>the computing system</nc> comprising <nc>a processor</nc> coupled to <nc>a memory</nc> and <nc>operative</nc> to implement: <nc>a normalizer</nc> for receiving and parsing <nc>the new concept descriptions</nc> so as to transform <nc>them</nc> into <nc>normalized ones</nc> and output <nc>them</nc>, wherein said <nc>normalized concept description</nc> includes <nc>one or more description</nc> <nc>part(s</nc>) having <nc>kernel concepts</nc>, said <nc>descriptions parts</nc> can only contain <nc>the terms</nc> <nc>which</nc> can be identified in <nc>said existing ontology</nc>, and <nc>each</nc> of <nc>said kernel concept</nc> contains <nc>a headword</nc> and <nc>a zero or more property(s</nc>); and <nc>a new concept factory</nc> for, based on <nc>the normalized description</nc> of <nc>the new concept</nc>, identifying <nc>the kernel concepts</nc> in <nc>each normalized concept description part</nc>, and extracting <nc>the identified kernel concepts</nc>, <nc>related properties</nc>, and <nc>the relations</nc> among <nc>the kernel concepts</nc> for <nc>a user</nc> to create <nc>new concepts</nc> according to <nc>existing ontologies</nc>; wherein <nc>the new concept factory</nc> comprises <nc>a kernel concept identifier</nc> for receiving <nc>the normalized new concept description</nc>, identifying <nc>the kernel concepts</nc> in <nc>each normalized concept description part</nc>, and extracting <nc>the identified kernel concepts</nc>, <nc>the related properties</nc>, and <nc>the relations</nc> among <nc>the kernel concepts</nc>; wherein <nc>an existing ontology</nc> is viewed as <nc>a directed graph</nc> <nc>G</nc>, where <nc>concepts</nc> are <nc>nodes</nc> and <nc>a relationship</nc> between <nc>two concepts</nc> is <nc>a directed edge</nc>; given n <nc>concepts</nc>, let <nc>c</nc> <nc>i</nc> denote <nc>the i-th concept</nc>, and let <nc>d(c</nc> <nc>i</nc> , <nc>c j</nc> ) be <nc>the distance</nc> between <nc>the i-th concept</nc> c <nc>i</nc> and <nc>the j-th concept</nc> <nc>c j</nc> in <nc>the directed graph</nc> <nc>G</nc>, s(node i ) be <nc>the total number</nc> of <nc>related concepts</nc> <nc>that</nc> <nc>the i-th</nc> concept c <nc>i</nc> can reach in <nc>the directed graph</nc> <nc>G</nc>, s(c <nc>i</nc> ) be <nc>the number</nc> of <nc>concepts</nc> <nc>the concept</nc> c <nc>i</nc> can reach in <nc>the description part</nc> and ∑ j = 1 n , <nc>j ≠ i</nc> <nc> d</nc> ( <nc>c i</nc> , <nc>c j</nc> ) be <nc>the sum</nc> of <nc>distances</nc> between <nc>the concept</nc> c <nc>i</nc> and <nc>all other concepts</nc> in <nc>the concept description</nc>, <nc>the importance</nc> <nc>D i</nc> of <nc>the concept</nc> c <nc>i</nc> can be calculated by <nc>the kernel concept identifier</nc> according to <nc>the following formula</nc>; <nc>D</nc> <nc>i</nc> = def <nc></nc> s <nc></nc> ( <nc>c i</nc> ) ∑ j = 1 n , <nc>j ≠ i</nc> d <nc></nc> ( <nc>c i</nc> , <nc>c j</nc> ) and <nc>the kernel concept</nc> <nc>c k</nc> in <nc>the description</nc> is thus obtained, wherein <nc>k</nc> can be determined by <nc>the following formula</nc>:
<nc>(1≦ k≦n</nc> )^<nc>( D k =Max</nc>( <nc>D</nc> <nc>i|j=1,n</nc> )).
| 6
|
6. <nc>The system</nc> for creating <nc>new concepts</nc> in <nc>existing ontologies</nc> according to <nc>claim</nc> 1 , further comprising <nc>a new concept expression generator</nc> for imposing <nc>property restrictions</nc> on <nc>the kernel concepts</nc> identified by <nc>the kernel concept identifier</nc> and/or performing intersection/union/complement <nc>operations</nc> on <nc>them</nc>, so as to generate <nc>a new concept expression</nc>.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.