doc_id
string | appl_id
string | claim1
string | claim_number
int64 | claim_text
string |
|---|---|---|---|---|
7516468
|
09976793
|
1. <nc>A method</nc> for transmitting <nc>interactive television information</nc> over <nc>a television broadcast</nc>, comprising: compiling <nc>business data</nc> into <nc>a binary form</nc>, <nc>the business data</nc> comprising <nc>descriptions</nc> of <nc>products</nc>, wherein said <nc>business data</nc> is compiled for <nc>use</nc> by <nc>a set-top box</nc>; generating <nc>a script</nc> using <nc>a script authoring tool</nc>, wherein said <nc>compiled business data</nc> is processed according to <nc>said generated script</nc> independent from <nc>a further user interaction</nc>; and transmitting <nc>a stream</nc>, comprising <nc>the compiled business data</nc> and <nc>the script</nc>, to <nc>a receiver</nc> for generating <nc>video information</nc> for <nc>a user's television</nc>, wherein <nc>the receiver</nc> uses <nc>the script</nc> to access <nc>the compiled business data</nc> and generate <nc>a presentation</nc> of <nc>the products</nc> for <nc>the user</nc>.
| 27
|
27. <nc>The method</nc> as described in <nc>claim</nc> 1 further comprising: providing <nc>content download</nc> in <nc>response</nc> to said <nc>user</nc> interacting with <nc>said presentation</nc> of <nc>said products</nc>.
|
8121838
|
11401792
|
1. <nc>A method</nc> for prioritizing <nc>speech recognition results</nc> from <nc>a plurality</nc> of <nc>speech recognition tasks</nc> to <nc>a human transcriptionist</nc> for <nc>evaluation</nc>, <nc>the method</nc> comprising <nc>acts</nc> of: accessing <nc>logged information</nc> generated during <nc>the plurality</nc> of <nc>speech recognition tasks</nc>, the accessing comprising accessing <nc>first logged information</nc> generated during <nc>a first speech recognition task</nc> and <nc>second logged information</nc> generated during <nc>a second speech recognition task</nc>, <nc>the first speech recognition task</nc> being performed on <nc>one or more first spoken utterances</nc> and producing <nc>a first recognized text</nc>, <nc>the second speech recognition task</nc> being performed on <nc>one or more second spoken utterances</nc> different from <nc>the one or more first spoken utterances</nc> and producing <nc>a second recognized text</nc> different from <nc>the first recognized text</nc>; associating <nc>a first accuracy rating</nc> with <nc>at least one portion</nc> of <nc>the first recognized text</nc> based at least in <nc>part</nc> on <nc>the first logged information</nc>, wherein <nc>the at least one portion</nc> of <nc>the first recognized text</nc> comprises <nc>a recognized phrase output</nc> by <nc>the first speech recognition task</nc> based on <nc>the one or more first spoken utterances</nc>, and wherein <nc>the first accuracy rating</nc> associated with <nc>the at least one portion</nc> of <nc>the first recognized text</nc> is based at least in <nc>part</nc> on <nc>at least one item</nc> of <nc>information</nc> <nc>that</nc> relates to <nc>the recognized phrase</nc> but is independent of <nc>a confidence score</nc> associated with <nc>the recognized phrase</nc> and independent of <nc>confidence scores</nc> associated with <nc>a plurality</nc> of <nc>phrases</nc> in <nc>an N-best match output</nc> by <nc>the first speech recognition task</nc>, <nc>the plurality</nc> of <nc>phrases</nc> in <nc>the N-best match</nc> being different from <nc>the recognized phrase</nc>; associating <nc>a second accuracy rating</nc> with <nc>at least one portion</nc> of <nc>the second recognized text</nc> based at least in <nc>part</nc> on <nc>the second logged information</nc>; and presenting <nc>the at least one portion</nc> of <nc>the first recognized text</nc> and <nc>the at least one portion</nc> of <nc>the second recognized text</nc> to <nc>the human transcriptionist</nc> for <nc>evaluation</nc>, wherein <nc>the at least one portion</nc> of <nc>the first recognized text</nc> is presented in <nc>such a manner</nc> as to be dissociated from <nc>at least one other portion</nc> of <nc>the first recognized text</nc>, and wherein <nc>at least one visual indication</nc> is provided to reflect <nc>a priority</nc> between <nc>the at least one portion</nc> of <nc>the first recognized text</nc> and <nc>the at least one portion</nc> of <nc>the second recognized text</nc>, <nc>the priority</nc> being determined based at least partially on <nc>the first and second accuracy ratings</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: providing <nc>an accuracy threshold</nc>; identifying <nc>a portion</nc> of <nc>a recognized text</nc> as being validated when <nc>a corresponding accuracy rating</nc> is above <nc>the accuracy threshold</nc>; and identifying <nc>a portion</nc> of <nc>a recognized text</nc> as being invalidated when <nc>a corresponding accuracy rating</nc> is below <nc>the accuracy threshold</nc>.
|
8867838
|
13612958
|
1. <nc>A computer implementable method</nc> for processing <nc>an electronic document</nc> comprising <nc>a plurality</nc> of <nc>text fields</nc> and <nc>a text label</nc> associated with <nc>each</nc> of <nc>the plurality</nc> of <nc>text fields</nc>, <nc>the computer implementable method</nc> comprising: extracting <nc>the plurality</nc> of <nc>text fields</nc> from <nc>the electronic document</nc>; grouping <nc>the plurality</nc> of <nc>extracted text fields</nc> to generate <nc>a plurality</nc> of <nc>groups</nc>; labeling <nc>the plurality</nc> of <nc>groups</nc> based on <nc>a first pre-defined criteria</nc> to generate <nc>a plurality</nc> of <nc>labeled groups</nc>; distributing <nc>the plurality</nc> of <nc>labeled groups</nc> in <nc>a plurality</nc> of <nc>queues</nc> based on <nc>a second pre-defined criteria</nc>; and transmitting <nc>the plurality</nc> of <nc>labeled groups</nc> from <nc>the plurality</nc> of <nc>queues</nc> to <nc>one or more crowdworkers</nc> based on <nc>a third pre-defined criteria</nc>.
| 19
|
19. <nc>The computer implementable method</nc> of <nc>claim</nc> 1 further comprising receiving <nc>one or more electronic documents</nc> from <nc>a database</nc>.
|
8738365
|
13616924
|
1. A computer readable storage device storing <nc>a program</nc> of <nc>instructions</nc> executable by <nc>a machine</nc> to perform <nc>a method</nc> of <nc>diffusing evidence</nc> among <nc>candidate answers</nc> during <nc>question</nc> answering, comprising: identifying, by <nc>a processor</nc>, <nc>a relationship</nc> between <nc>a first candidate answer</nc> and <nc>a second candidate answer</nc>, wherein <nc>the candidate answers</nc> are generated by <nc>a question-answering computer process</nc> for answering <nc>a question</nc>, <nc>the candidate answers</nc> have associated <nc>supporting evidence</nc>, and <nc>the candidate answers</nc> have associated <nc>confidence scores</nc>; determining whether to transfer <nc>the associated supporting evidence</nc> from <nc>the first candidate answer</nc> to <nc>the second candidate answer</nc>, or to transfer <nc>the associated supporting evidence</nc> from <nc>the second candidate answer</nc> to <nc>the first candidate answer</nc>, by analyzing how <nc>the question</nc> is posed and <nc>types</nc> of <nc>the first candidate answer</nc> and <nc>the second candidate answer</nc>; in <nc>response</nc> to determining to transfer <nc>the associated supporting evidence</nc> from <nc>the first candidate answer</nc> to <nc>the second candidate answer</nc>, transferring <nc>all</nc> or <nc>some</nc> of <nc>the evidence</nc> from <nc>the first candidate answer</nc> to <nc>the second candidate answer</nc> based on <nc>the identified relationship</nc>, and computing <nc>a new confidence score</nc> for <nc>the second candidate answer</nc> based on <nc>the transferred evidence</nc> and <nc>second candidate answer's existing evidence</nc>; in <nc>response</nc> to determining to transfer <nc>the associated supporting evidence</nc> from <nc>the second candidate answer</nc> to <nc>the first candidate answer</nc>, transferring <nc>all</nc> or <nc>some</nc> of <nc>the evidence</nc> from <nc>the second candidate answer</nc> to <nc>the first candidate answer</nc> based on <nc>the identified relationship</nc>, and computing <nc>a new confidence score</nc> for <nc>the first candidate answer</nc> based on <nc>the transferred evidence</nc> and <nc>first candidate answer's existing evidence</nc>.
| 4
|
4. <nc>The computer readable storage medium</nc> of <nc>claim</nc> 1 , wherein <nc>the transferring</nc> includes transferring <nc>feature scores</nc> across <nc>candidates</nc>.
|
9892469
|
13451351
|
1. <nc>A computer-implemented method</nc>, comprising: identifying, by <nc>one or more computing devices</nc>, <nc>a plurality</nc> of <nc>content component identifiers</nc> for <nc>an application</nc> running on <nc>a user computing device</nc>, <nc>each</nc> of <nc>the plurality</nc> of <nc>content component identifiers</nc> identifying a respective one of <nc>a plurality</nc> of <nc>separate content components</nc> presented by <nc>the application</nc>, wherein <nc>the plurality</nc> of <nc>separate content components</nc> include <nc>at least two different types</nc> of <nc>content</nc> including <nc>portions</nc> of <nc>text</nc>, <nc>media</nc>, <nc>images</nc>, or <nc>chat interfaces</nc> <nc>that</nc> are simultaneously presented within <nc>the application</nc>; receiving, by <nc>the one or more computing devices</nc> and through <nc>a first user interface</nc> of <nc>the application</nc>, <nc>an indication</nc> that <nc>a user</nc> recommended <nc>the application</nc> <nc>that</nc> is running on <nc>the user computing device</nc> and <nc>that</nc> presents <nc>the plurality</nc> of <nc>separate content components</nc> simultaneously; in <nc>response</nc> to receiving <nc>the indication</nc> that <nc>the user</nc> recommended <nc>the application</nc>: generating, by <nc>the one or more computing devices</nc> and in <nc>response</nc> to receiving <nc>the indication</nc>, <nc>a recommendation intent query</nc> in <nc>a second user interface</nc>, wherein <nc>the generated recommendation intent query</nc>: presents <nc>a list</nc> of <nc>the plurality</nc> of <nc>content component identifiers</nc> of <nc>the plurality</nc> of <nc>separate content components</nc> <nc>that</nc> were simultaneously presented by <nc>the application</nc> when <nc>the user</nc> recommended <nc>the application</nc>; and enables <nc>the user</nc> to choose, from <nc>the list</nc> of <nc>the plurality</nc> of <nc>content component identifiers</nc>, <nc>a given content component identifier</nc> <nc>that</nc> identifies <nc>a given content component</nc> <nc>that</nc> contributed to <nc>the user recommendation</nc> of <nc>the application</nc>; and transmitting <nc>the recommendation intent query</nc> for <nc>presentation</nc> at <nc>the user computing device</nc>; receiving, in <nc>response</nc> to <nc>presentation</nc> of <nc>the recommendation intent query</nc> and from <nc>the user computing device</nc> <nc>that</nc> presents <nc>the second user interface</nc>, <nc>an indication</nc> of <nc>user selection</nc> of at least one of <nc>the plurality</nc> of <nc>content component identifiers</nc> from <nc>the list</nc> of <nc>the plurality</nc> of <nc>content component identifiers</nc> <nc>that</nc> were presented by <nc>the recommendation intent query</nc>; generating, by <nc>the one or more computing devices</nc> and based on <nc>the user recommendation</nc> of <nc>the application</nc> by <nc>the user</nc> and <nc>the user selection</nc> of the at least one of <nc>the plurality</nc> of <nc>content component identifiers</nc>, <nc>a social annotation</nc> <nc>that</nc> communicates <nc>the recommendation</nc> of <nc>the application</nc> and <nc>any</nc> of <nc>the plurality</nc> of <nc>separate content components</nc> of <nc>the application</nc> <nc>that</nc> <nc>the user</nc> selected, from <nc>the list</nc> of <nc>the plurality</nc> of <nc>content component identifiers</nc>, as having contributed to <nc>the user recommendation</nc> of <nc>the application</nc>; and serving, by <nc>the one or more computing devices</nc> and via <nc>a network</nc>, <nc>the at least one social annotation</nc> to <nc>a second user computing device</nc> in <nc>a format</nc> suitable for <nc>presentation</nc> on <nc>the second user computing device</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the at least one social annotation</nc> served to <nc>the second user computing device</nc> identifies a specific one of <nc>the plurality</nc> of <nc>separate content components</nc> of <nc>the application</nc> <nc>that</nc> contributed to <nc>the recommendation</nc> of <nc>the application</nc> by <nc>the user</nc>.
|
9002961
|
13712416
|
1. A method of opening <nc>a communications channel</nc> between a first and <nc>a second person</nc>, comprising <nc>the steps</nc> of: receiving <nc>a date</nc> of <nc>birth</nc> from said first and said <nc>second person</nc>; <nc>polling</nc> said first and said <nc>second person</nc>, and receiving <nc>answers</nc> for said first and said <nc>second person</nc>; determining <nc>a Myers-Briggs Type Indicator</nc> (<nc>MBTI</nc>) for said first and said <nc>second person</nc> based on <nc>said answers</nc>; determining <nc>a Chinese zodiac sign</nc> corresponding to said first and said <nc>second person</nc>, based on said <nc>received date</nc> of <nc>birth</nc>; determining <nc>a Western astrological sign</nc> corresponding to said first and said <nc>second person</nc>, based on said <nc>received date</nc> of <nc>birth</nc>; receiving <nc>at least one attribute</nc> characteristic of said first and said <nc>second person</nc>, and receiving <nc>at least one limiting attribute</nc> required of <nc>a person</nc> from said first and said <nc>second person</nc>; exhibiting to said <nc>first person</nc>, <nc>information</nc> about said <nc>second person</nc> if <nc>a weighted ranking</nc> of <nc>said MBTI</nc>, said <nc>Chinese zodiac</nc>, and said <nc>Western astrological sign</nc> are determined to be above <nc>a compatibility threshold</nc>, and said <nc>second person</nc> comprises said <nc>at least one limiting attributes</nc> required of <nc>said first person</nc>; and providing <nc>a communication channel</nc> between said first and said <nc>second person</nc>, based on <nc>an expressed desire</nc> to communicate from said <nc>first person</nc> to said <nc>second person</nc>.
| 10
|
10. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>individuals</nc> are of <nc>the same type</nc> according to <nc>said MBTI</nc>, and <nc>a weighted average</nc> of <nc>compatibility</nc> between <nc>a western astrology</nc> and <nc>Chinese zodiac</nc> is low compared to <nc>a predefined threshold</nc>, then <nc>a match</nc> is suggested based on said <nc>MBTI compatibility rating</nc>.
|
8725723
|
12677454
|
1. <nc>A method</nc> for searching for <nc>a related term</nc> having <nc>rapidly increasing popularity</nc>, <nc>the method</nc> comprising: analyzing <nc>a search log</nc> and extracting <nc>a daily search frequency</nc> for <nc>search terms</nc> in <nc>the search log</nc>; comparing <nc>peaks</nc> of <nc>the daily search frequency</nc> extracted for <nc>the search terms</nc> in <nc>a period</nc>; analyzing <nc>relevance</nc> between <nc>candidate search terms</nc> for <nc>which</nc> <nc>the peaks</nc> have occurred together in <nc>the period</nc> with at least one of <nc>the search terms</nc> as <nc>a result</nc> of <nc>the comparison</nc>, and filtering out <nc>a candidate search term</nc> of <nc>the candidate search terms</nc> having <nc>no relevance</nc>; and wherein <nc>the extracting</nc> of <nc>the daily search frequency</nc> for <nc>search terms</nc> in <nc>the search log</nc> includes, extracting <nc>the daily search frequency</nc> for <nc>the search terms</nc> in <nc>the period</nc>, analyzing <nc>the daily search frequency</nc> for <nc>the search terms</nc>, extracting <nc>at least one search term</nc> having <nc>a search frequency</nc> <nc>that</nc> increases more rapidly than <nc>an increase reference value</nc> and decreases more rapidly than <nc>a decrease reference value</nc>, and extracting <nc>time information</nc> when <nc>a peak</nc> of <nc>the at least one search term</nc> has occurred.
| 9
|
9. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: recording and maintaining <nc>the search log</nc> comprising <nc>the daily search frequency</nc> in <nc>a database</nc>, wherein <nc>the analyzing</nc> of <nc>the search log</nc> and extracting of <nc>the daily search frequency</nc> for <nc>the search terms</nc> comprises analyzing <nc>the search log</nc> with <nc>reference</nc> to <nc>the database</nc> and extracting <nc>the daily search frequency</nc> for <nc>the search terms</nc>.
|
9442924
|
14974768
|
1. <nc>A method</nc> comprising: based at least in <nc>part</nc> on <nc>a user interaction</nc> with <nc>a game application</nc> <nc>that</nc> is executed on <nc>a client device</nc>, identifying <nc>a text string</nc> in <nc>the game application</nc>; receiving, from <nc>a server</nc> separate from <nc>the client device</nc>, <nc>a translation</nc> of <nc>the identified text string</nc>, <nc>the translation</nc> of <nc>the identified text string</nc> including <nc>a token</nc>; in <nc>an automated operation</nc> performed using <nc>one or more computer processor devices</nc>, substituting <nc>the token</nc> with <nc>a word</nc> in <nc>accordance</nc> with <nc>the user interaction</nc> with <nc>the game application</nc>, thereby providing <nc>an updated translation</nc> of <nc>the text string</nc>; and causing <nc>display</nc> of <nc>the updated translation</nc> of <nc>the text string</nc> on <nc>the client device</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , wherein: <nc>the translation</nc> of <nc>the identified text string</nc> is also <nc>a translation</nc> of <nc>a previous text string</nc>; and <nc>the method</nc> further comprises determining that <nc>the identified text string</nc> and <nc>the previous text string</nc> are identical.
|
9740735
|
11936098
|
1. <nc>A computer system</nc>, comprising: <nc>a distributed computer cluster</nc>; <nc>one or more processors</nc>; and <nc>one or more computer readable storage media</nc> having stored thereon computer-executable instructions <nc>that</nc> are executable by <nc>the one or more processors</nc> to cause <nc>the computer system</nc> to generate <nc>parallel-processing queries</nc>, <nc>the computer-executable instructions</nc> including <nc>instructions</nc> <nc>that</nc> are executable to cause <nc>the computer system</nc> to perform <nc>at least the following</nc>: create <nc>a structured query</nc> according to <nc>a structured query language</nc>, <nc>the structured query</nc> being created for <nc>execution</nc> in <nc>parallel</nc> across <nc>the distributed computer cluster</nc>; and; receive <nc>programming language syntax</nc> <nc>that</nc> comprises <nc>a functions</nc>; insert <nc>the received programming language syntax</nc> into <nc>the structured query</nc> of <nc>the structured query language</nc> such that <nc>a resulting query</nc> includes <nc>a structured query language statement</nc> in <nc>combination</nc> with <nc>programming language code</nc> <nc>that</nc> specifies <nc>both an aggregation</nc> and <nc>an operation</nc>; insert <nc>a keyword</nc> into <nc>the resulting query</nc> <nc>that</nc> defines <nc>an object type</nc> in <nc>the programming language code</nc>; compile <nc>the programming language code</nc>, wherein compiling <nc>the programming language code</nc> is performed based on <nc>the object type</nc> defined by <nc>the inserted keyword</nc>; and execute <nc>the resulting query</nc>, including <nc>both the structured query language</nc> and <nc>the programming language code</nc>, in <nc>a distributed manner</nc> on <nc>the distributed computer cluster</nc>.
| 5
|
5. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the syntax</nc> is employed with <nc>an aggregator</nc>.
|
9679079
|
14415689
|
1. <nc>A computer-implemented method</nc> executable by <nc>a server</nc>, <nc>the method</nc> comprising: receiving, from <nc>a client device</nc>, <nc>a first search query</nc> and <nc>a first search result request</nc> in <nc>respect</nc> of <nc>the first search query</nc>, <nc>the first search query</nc> including <nc>at least one first query search term</nc>; sending, to <nc>the client device</nc>, first search results of <nc>a first search</nc> conducted using <nc>the first search query</nc>; causing <nc>the client device</nc> to generate <nc>a search engine result page</nc> (<nc>SERP</nc>) comprising <nc>the first search results</nc>, <nc>the SERP</nc> further comprising (<nc>i</nc>) <nc>a search field</nc> with <nc>the at least one first query search term</nc> contained therein, and (ii) <nc>a unique uniform resource locator</nc> (<nc>URL</nc>) comprising <nc>an indication</nc> of <nc>the at least one first query search term</nc>; responsive to determining that <nc>a user</nc> of <nc>the client device</nc> has deleted <nc>the at least one first query search term</nc> from <nc>the search field</nc>, and is entering <nc>a second search query</nc> comprising <nc>at least one second query search term</nc> in <nc>the search field</nc> of <nc>the SERP</nc> containing <nc>the first search results</nc>: receiving, from <nc>the client device</nc>, elements of <nc>the second search query</nc>, the elements of <nc>the second search query</nc> including (<nc>i</nc>) <nc>at least the at least one first query search term</nc> obtained from <nc>the URL</nc> of <nc>the SERP</nc> without <nc>the server</nc> retrieving <nc>the at least one first query search term</nc> from <nc>stored data</nc> of <nc>the server</nc>, and <nc>(ii</nc>) <nc>the at least one second query search term</nc>; and prior to <nc>the server</nc> having received <nc>a second search result request</nc> from <nc>the client device</nc> in <nc>respect</nc> of <nc>the second search query</nc>, sending to <nc>the client device</nc>, at least one of (<nc>i</nc>) <nc>at least one search query suggestion</nc> based on, at least in <nc>part</nc>, <nc>the elements</nc> of <nc>the second search query</nc>, and (ii) second search results of <nc>a second search</nc> conducted using <nc>the at least one search query suggestion</nc>.
| 6
|
6. <nc>A computer-implemented method</nc> as recited in <nc>claim</nc> 1 , further comprising: receiving, from <nc>the client device</nc> via <nc>the at least one server</nc>, <nc>the first search result request</nc>; and causing, via <nc>the at least one server</nc>, <nc>the first search</nc> to be conducted to yield <nc>the first search results</nc>.
|
8019769
|
12016391
|
1. <nc>A method</nc> for determining <nc>valid citation patterns</nc> in <nc>text</nc> within <nc>an electronic document</nc> using <nc>a processor</nc>, <nc>the method</nc> comprising: accessing, from <nc>a memory</nc>, at least one citation pattern, wherein <nc>each citation pattern</nc> includes <nc>a set</nc> of <nc>citation components</nc> <nc>which</nc> together define <nc>a predetermined pattern</nc> of <nc>citation components</nc>, <nc>each citation component</nc> associated with <nc>a set</nc> of <nc>citation component criteria</nc>; comparing <nc>text</nc> in <nc>the electronic document</nc> with <nc>the predetermined pattern</nc> of <nc>the citation components</nc> corresponding to <nc>the at least one citation pattern</nc>; and determining <nc>valid citation patterns</nc> in <nc>the text</nc> by identifying <nc>the predetermined pattern</nc> of <nc>citation components</nc> corresponding to <nc>the at least one citation pattern</nc> in <nc>the text</nc>.
| 8
|
8. <nc>The method</nc> according to <nc>claim</nc> 1 , wherein <nc>the citation components</nc> comprise one or more of the following: <nc>a signal</nc>, <nc>a first party</nc>, <nc>a versus</nc>, <nc>a docket number</nc>, <nc>a skeletal citation</nc>, <nc>a reporter volume number</nc>, <nc>a reporter abbreviation</nc>, <nc>an initial page</nc>, <nc>a publishing service</nc>, <nc>an initial subdivision</nc>, <nc>an Internet citation</nc>, <nc>a pinpoint</nc>, <nc>a reporter table</nc>, <nc>a first parallel citation</nc>, <nc>a first parallel pinpoint</nc>, <nc>a second parallel citation</nc>, <nc>a second parallel pinpoint</nc>, <nc>an early American citation</nc>, or <nc>a court</nc> and <nc>date</nc> of <nc>decision</nc>.
|
7949524
|
11938802
|
1. A speech recognition apparatus comprising: <nc>a speech input unit</nc> configured to receive <nc>an input</nc> of <nc>a speech utterance</nc>; <nc>a keyword recognition unit</nc> configured to recognize <nc>a plurality</nc> of <nc>keywords</nc> included in <nc>the speech utterance</nc> as <nc>a recognition result</nc> using <nc>a processor</nc>; <nc>a presentation unit</nc> configured to present <nc>the recognition result</nc>; <nc>a correction input unit</nc> configured to receive <nc>a correction input</nc> for <nc>the recognition result</nc>; <nc>a correction unit</nc> configured to correct <nc>the recognition result</nc> based on <nc>the correction input</nc> to create <nc>a correction result</nc>; <nc>a dictionary generation unit</nc> configured to generate <nc>a standby-word dictionary</nc> as <nc>a union</nc> of <nc>recognition target vocabulary elements</nc> including <nc>the plurality</nc> of <nc>keywords</nc> corrected by <nc>the correction unit</nc> for recognizing <nc>the speech utterance</nc>; and <nc>a speech utterance recognition unit</nc> configured to recognize <nc>the speech utterance</nc> using <nc>the standby-word dictionary</nc>.
| 8
|
8. <nc>The speech recognition apparatus</nc> according <nc>to claim</nc> 1 wherein <nc>the presentation unit</nc> is configured to select <nc>keywords</nc> based on <nc>a probability</nc> of appearing within <nc>a recognition-target vocabulary element</nc> from among <nc>the recognition result</nc> and to present <nc>the selected keywords</nc>.
|
7917363
|
10545762
|
1. <nc>A process</nc> for estimating <nc>the speech recognition accuracy</nc> of <nc>a dialog system</nc>, including <nc>steps</nc> executed by <nc>a computer system</nc> comprising: generating <nc>a grammar</nc> from <nc>a plurality</nc> of <nc>example phrases</nc>; determining <nc>respective probabilities</nc> for correctly identifying <nc>words</nc> of <nc>an input phrase</nc> with <nc>corresponding words</nc> of <nc>said grammar</nc>; and generating <nc>a probability</nc> for correctly recognizing <nc>said input phrase</nc> by multiplying said <nc>respective probabilities</nc>.
| 2
|
2. <nc>A process</nc> as claimed in <nc>claim</nc> 1 , <nc>wherein said probabilities</nc> are <nc>probabilities</nc> of <nc>confusing words</nc> of <nc>said input phrase</nc> and <nc>words</nc> of <nc>said grammar</nc>.
|
8370147
|
13341621
|
1. <nc>A method</nc> for providing <nc>a natural language voice user interface</nc>, comprising: receiving <nc>a natural language utterance</nc> from <nc>an input device</nc> associated with <nc>a computing device</nc>, wherein <nc>the natural language utterance</nc> relates to <nc>navigation</nc>, and wherein <nc>the computing device</nc> is moving; determining <nc>a current location</nc> and <nc>direction</nc> of <nc>travel</nc> of <nc>the computing device</nc>; selecting, from among <nc>a plurality</nc> of <nc>sets</nc> of <nc>location-specific grammar information</nc>, <nc>a set</nc> of <nc>location-specific grammar information</nc> based on <nc>proximity</nc> between <nc>the current location</nc> and <nc>a location</nc> associated with <nc>the set</nc> of <nc>location-specific grammar information</nc> and based on whether <nc>the direction</nc> of <nc>travel</nc> of <nc>the computing device</nc> corresponds with <nc>movement</nc> towards <nc>the location</nc> associated with <nc>the set</nc> of <nc>location-specific grammar information</nc>; generating <nc>a recognition grammar</nc> with <nc>the set</nc> of <nc>location-specific grammar information</nc>; generating <nc>one or more interpretations</nc> of <nc>the natural language utterance</nc> using <nc>the recognition grammar</nc>; identifying, by <nc>a navigation agent</nc> executing on <nc>the computing device</nc>, <nc>one or more requests</nc> in <nc>the natural language utterance</nc> <nc>that</nc> relate to <nc>navigation</nc> from <nc>the one or more interpretations</nc> of <nc>the natural language utterance</nc>; and resolving, by <nc>the navigation agent</nc> executing on <nc>the computing device</nc>, <nc>the one or more requests</nc>.
| 17
|
17. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the plurality</nc> of <nc>geographic elements</nc> comprises <nc>a plurality</nc> of <nc>street elements</nc>, and wherein <nc>the</nc> generating <nc>the recognition grammar</nc> comprises removing at least one of <nc>the street elements</nc> from <nc>memory</nc>.
|
9424356
|
12964092
|
1. <nc>A computer-implemented method</nc> for updating <nc>a search index</nc> to identify <nc>documents</nc> with <nc>spiking interest</nc>, <nc>the method</nc> comprising: periodically receiving, at <nc>a first computing device</nc> having <nc>a processor</nc> and <nc>a memory</nc>, <nc>a history file</nc> from <nc>each</nc> of <nc>a plurality</nc> of <nc>second computing devices</nc>, <nc>the history file</nc> including <nc>a uniform resource identifier</nc> (<nc>URI</nc>) <nc>that</nc> has been accessed via a respective one of <nc>the plurality</nc> of <nc>second computing devices</nc> and <nc>metadata</nc> indicating <nc>a time</nc> of <nc>day</nc> when <nc>the URIs</nc> were accessed, wherein <nc>the plurality</nc> of <nc>second computing devices</nc> form <nc>a group</nc> of <nc>second computing devices</nc>; determining <nc>a number</nc> of <nc>instances</nc> of <nc>the URI</nc> in <nc>the history files</nc> for <nc>the group</nc> of <nc>second computing devices</nc>, wherein <nc>the number</nc> of <nc>instances</nc> occurs in <nc>a predetermined time period</nc> between <nc>a given start</nc> and stop <nc>time</nc>; calculating, automatically without <nc>user interaction</nc>, <nc>an average number</nc> of <nc>instances</nc> of <nc>the URI</nc> in <nc>the history files</nc> for <nc>the group</nc> of <nc>second computing devices</nc>, wherein <nc>the average number</nc> of <nc>instances</nc> occurs in <nc>the predetermined time period</nc> over <nc>a predetermined number</nc> of <nc>previous days</nc>, and wherein <nc>the predetermined time period</nc> is between <nc>the given start</nc> and stop <nc>time</nc> each day; determining that <nc>the number</nc> of <nc>instances</nc> of <nc>the URI</nc> in <nc>the history file</nc> is <nc>at least a predetermined amount</nc> greater than <nc>the average number</nc> of <nc>instances</nc> of <nc>the URI</nc> in <nc>the history files</nc> by, automatically without <nc>user interaction</nc>, comparing <nc>the number</nc> of <nc>instances</nc> of <nc>the URI</nc> for <nc>the group</nc> of <nc>second computing devices</nc> to <nc>the average number</nc> of <nc>instances</nc> of <nc>the URI</nc> over <nc>the predetermined number</nc> of <nc>days</nc> for <nc>said group</nc> of <nc>second computing devices</nc>; identifying <nc>the URI</nc> as <nc>a spiking URI</nc> when <nc>the number</nc> of <nc>instances</nc> of <nc>the URI</nc> is <nc>at least a predetermined amount</nc> greater than <nc>the average number</nc> of <nc>instances</nc> of <nc>the URI</nc> occurring over <nc>the predetermined number</nc> of <nc>days</nc>; and updating <nc>a search index</nc> to indicate <nc>the spiking URI</nc>, <nc>the search index</nc> being useable by <nc>a search engine</nc> to identify <nc>search results</nc> for <nc>a search query</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>each</nc> of <nc>the second computing devices</nc> includes <nc>a browser application</nc> <nc>that</nc> generates <nc>the history file</nc> and communicates <nc>the history file</nc> to <nc>the first computing device</nc>.
|
7958066
|
12262862
|
1. <nc>A method</nc> of helping <nc>a user</nc> make <nc>a decision</nc> through <nc>the use</nc> of <nc>a machine learning facility</nc>, comprising: creating <nc>a profile</nc> for <nc>the user</nc> through <nc>a sequence</nc> of <nc>questions</nc> presented from <nc>the machine learning facility</nc> to <nc>the user</nc> during <nc>a registration process</nc>; receiving <nc>an initial question</nc> at <nc>the machine learning facility</nc> from <nc>the user</nc>; providing <nc>the user</nc> with <nc>a dialog</nc> consisting of <nc>questions</nc> from <nc>the machine learning facility</nc> and <nc>answers</nc> provided by <nc>the user</nc>, wherein at least one of <nc>the questions</nc> from <nc>the machine learning facility</nc> is selected based upon <nc>the profile</nc>; including <nc>at least one randomly selected question</nc> in <nc>the dialogue</nc>, wherein <nc>the at least one randomly selected question</nc> is used to train <nc>the machine learning facility</nc> upon reaching <nc>a decision</nc>; and providing <nc>the decision</nc> to <nc>the user</nc> from <nc>the machine learning facility</nc>, wherein <nc>the decision</nc> is <nc>a single answer</nc> to <nc>the initial question</nc> from <nc>the user</nc> based on <nc>the dialog</nc> and <nc>the profile</nc>.
| 17
|
17. <nc>The method</nc> of <nc>claim</nc> 1 wherein <nc>the at least one randomly selected question</nc> is randomly selected from <nc>a plurality</nc> of <nc>suggested questions</nc> provided by <nc>users</nc> after <nc>completion</nc> of <nc>previous instances</nc> of <nc>a user dialogue</nc> related to <nc>the initial question</nc>.
|
9282076
|
13905792
|
1. <nc>A method</nc> for identifying <nc>questionable content</nc> in <nc>a proposed communication</nc> from <nc>a sender</nc>, wherein <nc>the proposed communication</nc> comprises <nc>initial content</nc>, <nc>the method</nc> comprising: receiving by <nc>a processor</nc> <nc>the initial content</nc>; receiving by <nc>the processor</nc> <nc>an identification</nc> of <nc>at least one proposed recipient</nc>; receiving, by <nc>the processor</nc>, private information associated with <nc>the proposed recipient</nc> from <nc>a source</nc> of <nc>private information</nc> about <nc>the proposed recipient</nc>; receiving, by <nc>the processor</nc>, public information associated with <nc>the proposed recipient</nc> from <nc>a source</nc> of <nc>public information</nc> about <nc>the proposed recipient</nc>; identifying by <nc>the processor</nc>, based upon <nc>the received initial content</nc>, the received private information and the received public information, <nc>at least a portion</nc> of <nc>the initial content</nc> as <nc>the questionable content</nc>; and indicating by <nc>the processor</nc>, to <nc>the sender</nc>, <nc>the identified questionable content</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the identification</nc> of <nc>the proposed recipient</nc> comprises one of: <nc>(a) a single email recipient</nc>; (b) <nc>an email distribution list</nc>; <nc>(c) a single instant message recipient</nc>; (<nc>d) a group instant chat list</nc>; <nc>(e) a social network distribution list</nc>; (f) a social network friends list; or <nc>(g</nc>) <nc>any combination</nc> thereof.
|
8015196
|
12155920
|
1. <nc>A method</nc>, implemented at least in <nc>part</nc> by <nc>a computer</nc>, of searching for <nc>geographic information</nc>, <nc>the method</nc> comprising <nc>the steps</nc> of: loading <nc>a dictionary</nc> of <nc>geographic feature names</nc> and <nc>geographic feature types</nc> into a memory communicably connected with <nc>the computer</nc>, wherein <nc>the loading</nc> is performed by <nc>the computer</nc>; <nc>loading locations</nc> associated with <nc>the geographic feature names</nc> into <nc>the memory</nc>, wherein <nc>the loading</nc> is performed by <nc>the computer</nc>; creating <nc>phonetic representations</nc> of <nc>the geographic feature names</nc> by normalizing <nc>each</nc> of <nc>the geographic feature names</nc>, wherein <nc>the creating</nc> is performed by <nc>the computer</nc>; storing <nc>the normalized geographic feature names</nc> in <nc>the memory</nc>; receiving, at <nc>the computer</nc>, <nc>a query</nc> for searching for <nc>geographic information</nc>, <nc>the query</nc> including <nc>a geographic name</nc>; creating <nc>a phonetic representation</nc> of <nc>the geographic name</nc> by normalizing <nc>the geographic name</nc> included in <nc>the query</nc>, wherein <nc>the creating</nc> is performed by <nc>the computer</nc>; searching <nc>the memory</nc> for <nc>a candidate name</nc> to be compared with <nc>the geographic name</nc> included in <nc>the query</nc>, <nc>the candidate name</nc> being one of <nc>the normalized geographic feature names</nc>; calculating <nc>a value</nc> of <nc>proximity</nc> between <nc>the candidate name</nc> and <nc>the geographic name</nc> included in <nc>the query</nc>, wherein <nc>the calculating</nc> is performed by <nc>the computer</nc>; and if <nc>the value</nc> of <nc>proximity</nc> is equal or larger than <nc>a predetermined value</nc>, outputting <nc>a geographic feature name</nc> corresponding to <nc>the candidate name</nc> and <nc>a location</nc> associated with <nc>the geographic feature name</nc> corresponding to <nc>the candidate name</nc>.
| 4
|
4. <nc>A method</nc> according to <nc>claim</nc> 1 , <nc>wherein the step</nc> of outputting <nc>comprises</nc> identifying <nc>the location</nc> associated with <nc>the geographic feature name</nc> corresponding to <nc>the candidate name</nc> on <nc>a map</nc>.
|
8775513
|
11948370
|
1. <nc>A method</nc> for messaging <nc>integration</nc> of <nc>a business object</nc> comprising: embedding <nc>a business object</nc> in <nc>message text</nc> in <nc>a messaging session</nc> provided by <nc>a messenger</nc>; applying <nc>an action</nc> to <nc>the business object</nc> from within <nc>the message session</nc> of <nc>the messenger</nc>; identifying <nc>a pronoun</nc> in <nc>the message text</nc> referencing <nc>the business object</nc>; and, visually distinguishing <nc>the identified pronoun</nc> in <nc>the message text</nc> to draw <nc>a correlation</nc> between <nc>the business object</nc> and <nc>the pronoun</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the messenger</nc> is <nc>an instant messenger</nc>, <nc>the message text</nc> is <nc>chat text</nc> and <nc>the messaging session</nc> is <nc>a chat session</nc>.
|
8996622
|
12241731
|
1. <nc>A method</nc>, comprising: generating by <nc>a network device</nc> <nc>one or more graphs</nc> using <nc>data</nc> obtained from <nc>a query log</nc>, <nc>the one or more graphs</nc> including <nc>an anticlick graph</nc>, wherein <nc>the anticlick graph</nc> represents <nc>information</nc> pertaining to <nc>documents</nc> in <nc>previously provided search results</nc> that, according to <nc>the data</nc> obtained from <nc>the query log</nc>, have not been clicked by <nc>a user</nc> <nc>that</nc> submitted <nc>a corresponding search query</nc> and does not represent <nc>information</nc> pertaining to <nc>documents</nc> in <nc>the previously provided search results</nc> that, according to <nc>the data</nc> obtained from <nc>the query log</nc>, have been clicked by <nc>the user</nc> <nc>that</nc> submitted <nc>the corresponding search query</nc>, wherein <nc>the anticlick graph</nc> includes <nc>one or more nodes</nc> representing or corresponding to <nc>documents</nc> that, according to <nc>the data</nc> obtained from <nc>the query log</nc>, have not been clicked by <nc>the user</nc> <nc>that</nc> submitted <nc>the corresponding search query</nc>; ascertaining by <nc>the network device values</nc> of <nc>one or more syntactic features</nc> of <nc>the one or more graphs</nc>; determining by <nc>the network device values</nc> of <nc>one or more semantic features</nc> of <nc>the one or more graphs</nc> by propagating <nc>categories</nc> from <nc>a web directory</nc> among <nc>nodes</nc> in <nc>each</nc> of <nc>the one or more graphs</nc>; and detecting by <nc>the network device spam hosts</nc> based upon <nc>the values</nc> of <nc>the syntactic features</nc> and <nc>the semantic features</nc>; wherein <nc>the anti-click graph</nc> includes <nc>a host-based graph</nc> or <nc>a document-based graph</nc>, wherein <nc>the nodes</nc> of <nc>the host-based graph</nc> includes <nc>one or more host nodes</nc> representing <nc>hosts</nc> corresponding to <nc>the documents</nc> <nc>that</nc>, according to <nc>the data</nc> obtained from <nc>the query log</nc>, have not been clicked by <nc>the user</nc> <nc>that</nc> submitted <nc>the corresponding search query</nc>, and wherein <nc>the nodes</nc> of <nc>the document-based graph</nc> includes <nc>one or more document nodes</nc> representing <nc>the documents</nc> <nc>that</nc>, according to <nc>the data</nc> obtained from <nc>the query log</nc>, have not been clicked by <nc>the user</nc> <nc>that</nc> submitted <nc>the corresponding search query</nc>.
| 12
|
12. <nc>The method</nc> as recited in <nc>claim</nc> 1 , wherein determining <nc>values</nc> of <nc>the one or more semantic features</nc> of <nc>the one or more graphs</nc> further comprises: categorizing by <nc>the network device</nc> <nc>a subset</nc> of <nc>hosts</nc> in <nc>the anti-click graph</nc> <nc>that</nc> can found in <nc>a web directory</nc> <nc>that</nc> includes <nc>a plurality</nc> of <nc>categories</nc> such that <nc>each</nc> of <nc>the subset</nc> of <nc>hosts</nc> is associated with one or more of <nc>the plurality</nc> of <nc>categories</nc>; and propagating by <nc>the network device</nc> the one or more of <nc>the plurality</nc> of <nc>categories</nc> to <nc>other host nodes</nc> and <nc>query nodes</nc> in <nc>the anti-click graph</nc> such that <nc>each node</nc> in <nc>the anti-click graph</nc> has <nc>an associated category tree</nc>.
|
9420355
|
14536441
|
1. <nc>A method</nc> comprising: receiving <nc>social object data</nc> including <nc>a plurality</nc> of <nc>metadata tags</nc>; receiving <nc>electronic program guide information</nc> including <nc>a plurality</nc> of <nc>television program identifiers</nc>; generating <nc>a graph data structure</nc> comprising <nc>a plurality</nc> of <nc>nodes</nc> and <nc>plurality</nc> of <nc>edges</nc>, <nc>each node</nc> representing <nc>a metadata tag</nc> of <nc>the plurality</nc> of <nc>metadata tags</nc> or <nc>a television program identifier</nc> of <nc>the plurality</nc> of <nc>television program identifiers</nc>, <nc>each edge</nc> connecting <nc>two nodes</nc>, <nc>each edge</nc> including <nc>a timestamp</nc> based on <nc>the social object data</nc>; receiving <nc>information</nc> about <nc>user-selected television shows</nc>; querying <nc>the graph data structure</nc> with <nc>a selected metadata tag</nc> of <nc>the plurality</nc> of <nc>metadata tags</nc> corresponding to <nc>a social object</nc> of <nc>the social object data</nc>; receiving <nc>a set</nc> of <nc>television program identifiers</nc> associated with <nc>the selected metadata tag</nc> by traversing, with <nc>a timestamp</nc> within <nc>a predetermined timeframe</nc>, <nc>at least a portion</nc> of <nc>the plurality</nc> of <nc>edges</nc> of <nc>the graph data structure</nc>; selecting <nc>a subset</nc> of <nc>the set</nc> of <nc>television program identifiers</nc> most closely related to <nc>the selected metadata tag</nc> by comparing <nc>the set</nc> of <nc>television program identifiers</nc> with <nc>the information</nc> about <nc>user-selected television shows</nc>; and responsive to one of <nc>the television program identifiers</nc> of <nc>the subset corresponding</nc> to at least one of <nc>the user-selected television shows</nc>, selecting <nc>the social object</nc> for <nc>removal</nc> from <nc>the social object data</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein generating <nc>a graph data structure</nc> further comprises generating <nc>edges</nc> of <nc>the plurality</nc> of <nc>edges</nc> based on <nc>data</nc> from <nc>the electronic program guide information</nc> indicating <nc>associations</nc> between <nc>metadata tags</nc> of <nc>the plurality</nc> of <nc>metadata tags</nc> and <nc>television program identifiers</nc> of <nc>the plurality</nc> of <nc>television program identifiers</nc>.
|
9589012
|
14815884
|
1. <nc>A computer-implemented method</nc>, comprising: receiving from <nc>a user</nc> <nc>a selection</nc> of <nc>an object</nc> among <nc>one or more objects</nc> included in <nc>a data model</nc>, <nc>the selection</nc> made through <nc>an object-selection interface</nc>; retrieving from <nc>computer memory</nc> <nc>a previously stored object definition</nc> <nc>that</nc> corresponds to <nc>the selected object</nc>, <nc>the previously stored object definition</nc> includes: <nc>an object query</nc> <nc>that</nc>, when executed, retrieves a set of <nc>time</nc> stamped events from <nc>a data store</nc> on <nc>a computing device</nc>, <nc>each event</nc> including <nc>a portion</nc> of <nc>raw machine data</nc> reflecting <nc>activity</nc> in <nc>an information technology environment</nc>; and <nc>an object schema</nc> identifying <nc>a set</nc> of <nc>one or more fields</nc>, <nc>each field</nc> defined by <nc>an extraction rule</nc> or <nc>regular expression</nc> <nc>that</nc> locates <nc>the field</nc> in <nc>the raw machine data</nc> and can be used to extract <nc>a field value</nc> from <nc>the field location</nc> from <nc>the raw machine data</nc> in <nc>each event</nc> in <nc>a subset</nc> of <nc>the set</nc> of <nc>time</nc> <nc>stamped events</nc>, <nc>each extraction rule</nc> or <nc>regular expression</nc> operating on <nc>the raw machine data</nc> in <nc>an event</nc> without modifying <nc>the event's raw machine data</nc>; and executing, against <nc>events</nc> in <nc>the data store</nc> <nc>that</nc> meet <nc>filtering criteria</nc> of <nc>the object query</nc>, <nc>a search query</nc> that references <nc>only field values</nc> <nc>that</nc> are extracted using <nc>the object schema</nc> and <nc>that</nc> produces <nc>a result</nc> based at least in <nc>part</nc> on <nc>the data</nc> reflecting <nc>the activity</nc> of <nc>the information technology environment</nc>.
| 22
|
22. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the search query</nc> is selected by <nc>a user</nc> from <nc>a displayed list</nc> of <nc>pre-defined queries</nc>.
|
8001465
|
10020909
|
1. <nc>A computer executable method</nc> for displaying <nc>elements</nc> of <nc>an information array</nc> within <nc>a predetermined two dimensional display space</nc>, wherein <nc>the predetermined two dimensional display space</nc> is divided into <nc>cells</nc> formed at <nc>intersections</nc> of <nc>columns</nc> and <nc>rows</nc>, <nc>the elements</nc> of <nc>the information array</nc> have <nc>corresponding cells</nc> for <nc>display</nc>, and at least two of <nc>said elements</nc> include <nc>text</nc>, said <nc>method</nc> comprising <nc>the steps</nc> of: (a) determining <nc>display space requirement</nc> (<nc>DSR</nc>) for displaying <nc>the elements</nc>; (b) moderating <nc>the DSR value</nc> of <nc>at least one element</nc> to determine <nc>its moderated display space requirement</nc> <nc>(ModDSR) value</nc>, wherein said <nc>moderating step</nc> comprises: <nc>(i</nc>) selecting <nc>an element</nc> <nc>whose DSR value</nc> is larger than <nc>the DSR value</nc> of <nc>at least one element</nc> in <nc>the column</nc> or <nc>row</nc> to <nc>which</nc> said <nc>element</nc> corresponds; and (ii) reducing <nc>the DSR value</nc> of <nc>the selected element</nc> such that <nc>the amount</nc> of <nc>reduction</nc> depends on <nc>the difference</nc> between <nc>the DSR value</nc> of <nc>said element</nc> and <nc>a value representative</nc> of <nc>the DSR values</nc> of <nc>the elements</nc> corresponding to <nc>the column</nc> or row to <nc>which</nc> said <nc>element</nc> corresponds; (c) allocating <nc>column widths</nc> and <nc>row heights</nc>, based on <nc>the ModDSR values</nc> or on <nc>values</nc> obtained by using <nc>the ModDSR values</nc>, such that <nc>the total width</nc> of <nc>all the columns</nc> and <nc>the total height</nc> of <nc>all the rows</nc> do not exceed <nc>the width</nc> and <nc>height</nc>, respectively, of <nc>the predetermined two dimensional display space</nc>; and (d) displaying <nc>the elements</nc> in <nc>the space</nc> allocated to <nc>the corresponding cells</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 wherein in <nc>step</nc> (a) <nc>the DSR</nc> of <nc>text elements</nc> is determined after abbreviating <nc>the text</nc>.
|
9203849
|
14097140
|
1. <nc>A method</nc> for detecting <nc>illegitimate links</nc> on <nc>a web page</nc>, <nc>the method</nc> comprising <nc>steps</nc> <nc>that</nc> include: receiving <nc>a web link</nc> <nc>that</nc> includes <nc>link text</nc> and <nc>a link address</nc>; generating <nc>normalized link text</nc> based on <nc>the link text</nc>, <nc>wherein one or more characters</nc> in <nc>the link text</nc> <nc>that</nc> are visually similar are represented by <nc>a single normalized character identifier</nc> in <nc>the normalized link text</nc>; determining whether <nc>the normalized link text</nc> is in <nc>a format</nc> of <nc>the link address</nc>; when (1) <nc>the normalized link text</nc> is not in <nc>the format</nc> of <nc>the link address</nc>, and (2) there is <nc>additional context text</nc> adjacent to <nc>the normalized link text</nc>: extending <nc>the normalized link text</nc> to include <nc>the additional context text</nc>, wherein <nc>the additional context text</nc> includes <nc>context text</nc> from <nc>a left side</nc> or <nc>a right side</nc> of <nc>the normalized link text</nc>, and repeating <nc>the steps</nc> of generating, <nc>determining</nc>, and <nc>responsive actions</nc>, wherein <nc>the normalized link text</nc> is generated based on <nc>the additional context text</nc>; and determining that <nc>the link text</nc> is safe when there is <nc>no additional context text</nc> adjacent to <nc>the normalized link text</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the one or more characters</nc> <nc>that</nc> are visually similar include <nc>characters</nc> <nc>that</nc> are likely to be confused with each other.
|
4448423
|
06419756
|
1. <nc>A board game apparatus</nc> comprising <nc>a playing board</nc> having <nc>a flat surface</nc>, to receive, thereon, playing <nc>pieces</nc> having <nc>a definite configuration</nc>, and adapted to be used by <nc>at least two players</nc> comprising: means on said <nc>flat surface dividing</nc> said <nc>surface</nc> into <nc>first and second sections</nc>, <nc>the surfaces</nc> of said <nc>first and second sections</nc> adapted to receive <nc>playing pieces</nc> placed in <nc>juxtaposition</nc> on <nc>said surfaces</nc> to form <nc>words</nc>, <nc>a plurality</nc> of <nc>sets</nc> of playing <nc>pieces</nc>, means disposed on <nc>each</nc> <nc>said set</nc> of playing <nc>pieces</nc> distinguishing <nc>each</nc> said set from <nc>other sets</nc>, means disposed on <nc>the top surface</nc> of <nc>certain playing pieces</nc> of <nc>each</nc> said <nc>set representing alphabetical characters</nc> selected from <nc>the group</nc> consisting of <nc>consonants</nc> and <nc>vowels</nc>, other means disposed on <nc>the top surface</nc> of <nc>certain playing pieces</nc> of <nc>each</nc> said <nc>set</nc> representing <nc>the absence</nc> of <nc>said alphabetical characters</nc>, other means disposed on <nc>certain playing pieces</nc> of <nc>each</nc> said set whereby <nc>the playing pieces</nc> of <nc>each</nc> said <nc>set</nc> are divided into <nc>two subgroups</nc> comprising <nc>a first group</nc> consisting of playing <nc>pieces</nc> having <nc>alphabetical consonants</nc> on <nc>the top surface</nc> and <nc>a second group</nc> consisting of playing <nc>pieces</nc> having <nc>alphabetical vowel characters</nc> or <nc>the absence</nc> of <nc>alphabetical characters</nc> on <nc>the top surface</nc>.
| 3
|
3. <nc>The board game apparatus</nc> of <nc>claim</nc> 1 further comprising <nc>a playing piece holder</nc> for holding <nc>each</nc> said set of playing <nc>pieces</nc> during <nc>the playing</nc> of <nc>the game</nc> comprising, means for holding <nc>a portion</nc> of <nc>the playing pieces</nc> in <nc>a top surface up position</nc> ready for <nc>view</nc> and play by <nc>the player</nc>, and <nc>other means</nc> for holding <nc>the remaining playing pieces</nc> in <nc>a top surface</nc> down <nc>position</nc> ready for <nc>selection</nc> by <nc>the player</nc>.
|
9484023
|
13773880
|
1. <nc>A method</nc>, comprising: converting <nc>a non-back-off language model</nc> to <nc>a back-off language model</nc> using <nc>a background language model</nc>, wherein <nc>the non-back-off language model</nc> assigns <nc>a probability</nc> to any fixed order <nc>n</nc>-gram without backing off to <nc>lower order n-gram probabilities</nc>, and wherein <nc>the background language model</nc> assigns <nc>a non-zero probability</nc> to n<nc>-grams</nc> assigned <nc>a zero probability</nc> by <nc>the non-back-off language model</nc>; and pruning <nc>the converted back-off language model</nc>; wherein <nc>the converted back-off language model</nc> assigns <nc>a probability</nc> to <nc>a given fixed order n-gram</nc> assigned <nc>a zero probability</nc> by <nc>the non-back-off language model</nc> and not assigned <nc>a fixed order probability</nc> by <nc>the background language model</nc> by backing off to <nc>a given lower order</nc> n<nc>-gram</nc> <nc>corresponding</nc> to <nc>the given fixed order</nc> n-<nc>gram</nc>; wherein <nc>the given fixed order n-gram</nc> comprises <nc>a given word</nc> and <nc>an associated history</nc>; wherein <nc>the</nc> given <nc>lower order</nc> n<nc>-gram</nc> comprises <nc>the given word</nc> and <nc>a truncated version</nc> of <nc>the associated history</nc>; wherein <nc>lower order</nc> n<nc>-grams</nc> originate from <nc>at least one lower order non-back-off language model</nc> after converting <nc>the lower order</nc> <nc>non-back-off language model</nc> to <nc>the back-off language model</nc> and <nc>highest order</nc> n<nc>-grams</nc> originate from <nc>an n-gram language model</nc>; and wherein <nc>the converted back-off language model</nc> is usable for decoding <nc>speech</nc>, and <nc>the converting and pruning steps</nc> are executed via <nc>a processor device</nc> configured to implement <nc>at least one speech decoder</nc> associated with <nc>an automatic speech recognition system</nc> configured to integrate <nc>the converted back-off language model</nc> into <nc>a decoding process</nc>.
| 16
|
16. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising normalizing <nc>back-off weights</nc> of <nc>the background language model</nc>.
|
8099397
|
12548217
|
1. <nc>A method</nc> for <nc>improved Portable Document Format (“PDF”) document archiving</nc>, <nc>the method</nc> comprising: scanning by <nc>use</nc> of <nc>a processor</nc>, <nc>a source Portable Document Format (“PDF”) document</nc> for <nc>a shared resource</nc>, <nc>the source PDF document</nc> comprising <nc>a plurality</nc> of <nc>records</nc>, <nc>the shared resource</nc> comprising <nc>a common resource</nc> referenced by <nc>way</nc> of <nc>a resource pointer</nc> associated with <nc>a record</nc> of <nc>the source PDF document</nc>; copying <nc>the shared resource</nc> to <nc>a resource group</nc> associated with <nc>the source PDF document</nc>; short-circuiting <nc>a link</nc> between <nc>content</nc> for <nc>the shared resource</nc> and <nc>the resource pointer</nc> in <nc>each record</nc> <nc>that</nc> points to <nc>the shared resource</nc>, wherein short-circuiting <nc>a link</nc> further comprises modifying <nc>the resource pointer</nc> to point to <nc>the copied shared resource</nc> in <nc>the resource group</nc> and wherein short-circuiting <nc>a link</nc> between <nc>the shared resource</nc> and <nc>the resource pointer</nc> further comprises removing <nc>content</nc> for <nc>the shared resource</nc> from <nc>the source PDF document</nc>; and extracting <nc>a record</nc> from <nc>the source PDF document</nc>, <nc>the extracted record void</nc> of <nc>the content</nc> for <nc>the shared resource</nc> in <nc>response</nc> to <nc>the short-circuited link</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising <nc>indexing index data</nc> in <nc>the source PDF document</nc> and storing <nc>the index data</nc> in <nc>a searchable repository</nc>.
|
9367434
|
14044467
|
1. <nc>A method</nc> comprising: receiving, by <nc>a processor</nc> of <nc>a system</nc>, <nc>a policy-based Extensible Markup Language (XML) workflow</nc> comprising <nc>multiple policy nodes</nc> and <nc>multiple condition nodes</nc>, <nc>the multiple policy nodes</nc> and <nc>multiple condition nodes</nc> accessible through <nc>a common input point</nc> in <nc>the first policy-based XML workflow</nc>; <nc>parsing</nc>, by <nc>the processor</nc>, <nc>the policy-based XML workflow</nc> to construct <nc>a policy control flow graph</nc> for <nc>the policy-based XML workflow</nc>; identifying, by <nc>the processor</nc>, multiple workflow subpaths in <nc>the policy control flow graph</nc>; identifying <nc>subpath constraints</nc> for traversing <nc>the workflow subpaths</nc> in <nc>the policy control flow graph</nc>; determining, by <nc>the processor</nc>, <nc>path constraints</nc> for <nc>a selected path</nc> in <nc>the policy control flow graph</nc> by: determining <nc>constituent subpaths</nc> from among <nc>the workflow subpaths</nc> <nc>that</nc> together traverse <nc>the selected path</nc> in <nc>the policy control flow graph</nc>; and determining <nc>the path constraints</nc> for <nc>the selected path</nc> by collecting <nc>those subpath constraints</nc> for traversing <nc>the constituent subpaths</nc> <nc>that</nc> traverse <nc>the selected path</nc>; generating, with <nc>the processor</nc>, <nc>a set</nc> of <nc>test inputs</nc> for <nc>the policy-based XML workflow</nc> responsive to <nc>the path constraints</nc> for <nc>the selected path</nc>, where <nc>the set</nc> of <nc>test inputs</nc>, when input into <nc>the policy-based XML workflow</nc>, cause <nc>the policy-based XML workflow</nc> to traverse <nc>the selected path</nc> in <nc>the policy control flow graph</nc>; and storing, by <nc>the processor</nc>, <nc>the generated set</nc> of <nc>test inputs</nc> in <nc>a memory</nc> of <nc>the system</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , where determining <nc>path constraints</nc> for <nc>selected path</nc> in <nc>the policy control flow graph</nc> comprises: identifying, in <nc>the policy control flow graph</nc>, a first node and a second node linked to <nc>the first node</nc>; determining <nc>a path condition</nc> for traversing from <nc>the first node</nc> to <nc>the second node</nc>; and appending <nc>the path condition</nc> to <nc>a previously determined path condition</nc> for reaching <nc>the first node</nc>.
|
10002301
|
15708485
|
1. <nc>A method</nc> for <nc>Arabic handwriting recognition</nc>, <nc>the method</nc> comprising: acquiring, an input image representative of <nc>a handwritten Arabic text</nc> from <nc>a user</nc>; partitioning, using <nc>processing circuitry</nc> of <nc>a server</nc>, <nc>the input image</nc> into <nc>a plurality</nc> of <nc>regions</nc>; determining, using <nc>the processing circuitry</nc>, <nc>a bag</nc> of <nc>features representation</nc> for <nc>each region</nc> of <nc>the plurality</nc> of <nc>regions</nc>; modeling, using <nc>the processing circuitry</nc>, <nc>each region</nc> independently by <nc>multi stream discrete Hidden Markov Model</nc> (<nc>HMM</nc>); and identifying, using <nc>processing circuitry</nc>, <nc>a recognized text</nc> based on <nc>the HMM models</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining <nc>a gradient magnitude</nc> and <nc>orientation</nc> by referencing <nc>a lookup table</nc> stored in <nc>a memory</nc> of <nc>the server</nc>.
|
9965232
|
15070277
|
1. <nc>A computer-implemented method</nc> for automatically verifying <nc>default printing selections</nc>, <nc>at least a portion</nc> of <nc>the method</nc> being performed by <nc>a computing device</nc> comprising <nc>at least one processor</nc>, <nc>the method</nc> comprising: receiving <nc>a digital printing request</nc> from <nc>a user</nc> of <nc>the computing device</nc> to print <nc>a document</nc>; identifying, in <nc>response</nc> to receiving <nc>the digital printing request</nc>, <nc>a default printer</nc> to <nc>which</nc> <nc>the computing device</nc> is configured to transmit <nc>instructions</nc> to print <nc>the document</nc>; identifying <nc>a policy</nc>, wherein: <nc>the policy</nc> comprises <nc>a printer-selection directive</nc> to: automatically print to <nc>the default printer</nc> without requesting <nc>a user verification</nc> whenever <nc>the default printer</nc> is within <nc>a designated area</nc> of <nc>the computing device</nc>; and verify <nc>a default printer selection</nc> prior to printing to <nc>the default printer</nc> whenever <nc>the default printer</nc> is not within <nc>the designated area</nc> of <nc>the computing device</nc>; and <nc>the policy</nc> further comprises <nc>a security policy directive</nc> to scan <nc>one or more portions</nc> of <nc>the document</nc> for <nc>sensitive information</nc> by scanning <nc>the one or more portions</nc> for at least one of <nc>a particular word</nc> and <nc>a particular phrase</nc>; determining that <nc>the default printer</nc> is not within <nc>the designated area</nc> of <nc>the computing device</nc>; performing <nc>a scan</nc> as directed by <nc>the security policy directive</nc> and <nc>determining</nc>, based on <nc>a result</nc> of <nc>the scan</nc>, that <nc>the document</nc> includes <nc>sensitive information</nc>; and based on <nc>the policy</nc>, and in <nc>response</nc> to determining <nc>both</nc> that <nc>the default printer</nc> is not within <nc>the designated area</nc> of <nc>the computing device</nc> and that <nc>the document</nc> includes <nc>the sensitive information</nc>, presenting <nc>the user</nc> with <nc>a verification prompt</nc> <nc>that</nc> cautions <nc>the user</nc> that <nc>the default printer</nc> is not within <nc>the designated area</nc> of <nc>the computing device</nc> and allows <nc>the user</nc> to: affirmatively select <nc>the default printer</nc> to print <nc>the document</nc>; or select <nc>a different printer</nc> to print <nc>the document</nc>.
| 5
|
5. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , further comprising: after presenting <nc>the user</nc> with <nc>the verification prompt</nc>, receiving <nc>digital input</nc> from <nc>the user</nc> <nc>that</nc> includes <nc>a selection</nc> of <nc>a different printer</nc> for printing <nc>the document</nc>; and in <nc>response</nc> to receiving <nc>the digital input</nc>, transmitting <nc>instructions</nc> to print <nc>the document</nc> to <nc>the different printer</nc> instead of transmitting <nc>the instructions</nc> to <nc>the default printer</nc>.
|
8468166
|
13050883
|
1. <nc>A method</nc> for <nc>processing queries</nc> comprising: executing <nc>a query</nc> to generate <nc>a relation</nc> defined by <nc>a model clause</nc> specified in <nc>said query</nc>, said <nc>model clause</nc> specifying: <nc>one or more dimension columns</nc> of <nc>said relation</nc>, and <nc>a rule</nc>, wherein said <nc>rule</nc> comprises: <nc>a left-side expression</nc>, <nc>wherein said left-side expression</nc> includes <nc>one or more existential predicates</nc>, wherein <nc>each</nc> of said <nc>one or more existential predicates</nc> evaluates to <nc>a Boolean value</nc> and refers to at least one of said <nc>one or more dimension columns</nc> of <nc>said relation</nc>; and <nc>a right-side expression</nc>; wherein executing said <nc>query</nc> comprises evaluating <nc>said rule</nc>, wherein evaluating said <nc>rule</nc> includes performing <nc>an UPSERT operation</nc> to insert <nc>a row</nc> into <nc>said relation</nc>; and wherein <nc>the method</nc> is performed by <nc>one or more computer systems</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein: said <nc>left-side expression</nc> further includes <nc>one or more qualified predicates</nc>, wherein <nc>each</nc> of said <nc>one or more qualified predicates</nc> generates <nc>a single value</nc>; wherein <nc>each</nc> of said <nc>one or more qualified predicates</nc> corresponds to <nc>one dimension column</nc> of said <nc>one or more dimension columns</nc>; said <nc>left-side expression</nc> defines <nc>a set</nc> of <nc>rows</nc> in <nc>said relation</nc>, wherein <nc>each row</nc> of <nc>said set</nc> of <nc>rows</nc> includes: <nc>each</nc> of said <nc>one or more dimension columns</nc> <nc>that</nc> corresponds to <nc>each</nc> of said <nc>one or more existential predicates</nc>; and <nc>each</nc> of said <nc>one or more dimension columns</nc> <nc>that</nc> corresponds to <nc>each</nc> of said <nc>one or more qualified predicates</nc>; and after evaluating <nc>said rule</nc>, said set of <nc>rows</nc> includes <nc>rows</nc> for <nc>all combinations</nc> of: <nc>each</nc> of said <nc>one or more dimension values</nc> for <nc>which</nc> <nc>each</nc> of said <nc>one or more existential predicates</nc> evaluates to <nc>a Boolean TRUE value</nc>; and <nc>each single value</nc> generated by <nc>each</nc> of said <nc>one or more qualified predicates</nc>.
|
8200683
|
12495663
|
1. <nc>A computer-implemented method</nc> comprising: processing, using <nc>a hardware processor</nc>, <nc>a search query</nc> specifying <nc>one or more keywords</nc> to identify <nc>listings</nc> satisfying <nc>the search query</nc>, <nc>each listing</nc> from <nc>the identified listings</nc> describing <nc>an item</nc> or <nc>service</nc> offered for <nc>sale</nc> and including <nc>a title</nc> and <nc>a price</nc>; organizing <nc>the identified listings</nc> based on <nc>a composite relevance score</nc> assigned to <nc>each listing</nc> from <nc>the identified listings</nc>, the composite relevance score derived as <nc>a sum</nc> of <nc>relevance scores</nc> respectively associated with <nc>each keyword</nc> from <nc>the one or more keywords</nc> appearing in <nc>a listing</nc> from <nc>the identified listings</nc>, <nc>each relevance score</nc> representing <nc>a customer desirability measure</nc> for <nc>a keyword</nc> in <nc>the listing</nc>, <nc>the customer desirability measure</nc> determined based on <nc>a calculated difference</nc> between <nc>a measure</nc> of <nc>customer demand</nc> and <nc>supply</nc> for <nc>the keyword</nc>, <nc>the measure</nc> of <nc>customer demand</nc> for <nc>the keyword</nc> is expressed as <nc>the percentage</nc> of <nc>listings</nc> <nc>that</nc> <nc>i</nc>) are selected by <nc>a user</nc> over <nc>a period</nc> of <nc>time</nc>, and <nc>ii</nc>) have <nc>titles</nc> including <nc>words</nc> <nc>that</nc> match <nc>keywords</nc> specified in <nc>the query</nc>, and <nc>the measure</nc> of <nc>supply</nc> for <nc>the keyword</nc> is expressed as <nc>the percentage</nc> of <nc>listings</nc> satisfying <nc>the search query</nc> <nc>that</nc> include <nc>the keyword</nc>; and displaying <nc>the organized listings</nc>.
| 4
|
4. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , wherein <nc>the measure</nc> of <nc>customer demand</nc> for <nc>the keyword</nc> in <nc>the title</nc> of <nc>the listing</nc> is expressed as <nc>the percentage</nc> of <nc>all search queries</nc> received over <nc>a period</nc> of <nc>time</nc> <nc>that</nc> include <nc>the keyword</nc>.
|
9652717
|
14576854
|
1. <nc>A method</nc>, comprising: receiving <nc>a case</nc>; generating <nc>a set</nc> of <nc>candidate answers</nc> for <nc>the case</nc> based on <nc>a corpus</nc> of <nc>information</nc>; excluding <nc>a first candidate answer</nc> from <nc>the set</nc> of <nc>candidate answers</nc> upon determining that <nc>a first attribute</nc> in <nc>the case</nc> precludes returning <nc>the first candidate answer</nc> as <nc>a valid response</nc> to <nc>the case</nc> based on <nc>a first rule</nc>, of <nc>a plurality</nc> of <nc>rules</nc> for <nc>processing supporting evidence</nc>; processing supporting <nc>evidence</nc> for <nc>the remaining candidate answers</nc> in <nc>the set</nc> of <nc>candidate answers</nc> by searching for <nc>items</nc> of <nc>evidence</nc> in <nc>the corpus</nc> of <nc>information</nc> <nc>that</nc> include <nc>passages</nc> supporting <nc>at least one candidate answer</nc> for <nc>each candidate answer</nc> in <nc>the set</nc> of <nc>candidate answers</nc>; and foregoing <nc>processing supporting evidence</nc> for <nc>the first candidate answer</nc> by refraining from searching for <nc>items</nc> of <nc>evidence</nc> in <nc>the corpus</nc> of <nc>information</nc> <nc>that</nc> include <nc>passages</nc> supporting <nc>the first candidate answer</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the set</nc> of <nc>candidate answers</nc> specifies <nc>candidate answers</nc> for <nc>which</nc> <nc>supporting evidence</nc> will be processed, wherein <nc>supporting evidence</nc> is not processed for <nc>the first candidate answer</nc>, wherein <nc>the first rule</nc> comprises <nc>a minimum confidence threshold</nc> for <nc>a confidence score</nc> of <nc>each</nc> of <nc>the candidate answers</nc>, wherein <nc>the attribute</nc> of <nc>the first candidate answer</nc> comprises <nc>the confidence score</nc> for <nc>the first candidate answer</nc>, wherein <nc>the first candidate answer</nc> is excluded from <nc>the set</nc> of <nc>candidate answers</nc> upon determining that <nc>the confidence score</nc> for <nc>the first candidate answer</nc> is below <nc>the minimum confidence threshold</nc>, wherein foregoing <nc>processing supporting evidence</nc> for <nc>the first candidate answer</nc> is based upon <nc>the determination</nc> that <nc>the confidence score</nc> for <nc>the first candidate answer</nc> is below <nc>the minimum confidence threshold</nc>.
|
7836394
|
11405501
|
1. <nc>A computer program product</nc>, comprising <nc>a computer-readable storage medium</nc> including <nc>computer-readable instructions</nc> embodied therein, that when executed by <nc>one or more processors</nc>, implement <nc>a method</nc> for <nc>the retrieval</nc>, <nc>analysis</nc> and <nc>display</nc> of <nc>electronically tagged financial data</nc>, <nc>the instructions</nc> comprising: <nc>an integrated file access component</nc>, <nc>analysis component</nc> and <nc>presentation component</nc> for accessing, analyzing and presenting <nc>electronically tagged financial data</nc> within <nc>the application</nc>; <nc>the file access component</nc> comprising: <nc>one or more user selection modules</nc> for selecting <nc>a plurality</nc> of <nc>files</nc>, <nc>the files</nc> including <nc>at least electronically tagged financial data</nc> <nc>which</nc> adhere to <nc>different taxonomies</nc> <nc>that</nc> are associated with <nc>the same type</nc> of <nc>content</nc>, <nc>the analysis component</nc> comprising: <nc>one or more analysis modules</nc> for calculating <nc>at least a first common analysis measure</nc> for <nc>comparison</nc> and <nc>analysis</nc> of <nc>electronically tagged financial data</nc> from <nc>a first file</nc> <nc>which</nc> adheres to <nc>a first taxonomy</nc> and <nc>corresponding electronically tagged financial data</nc> from <nc>a second file</nc> <nc>which</nc> adheres to <nc>a second taxonomy</nc> <nc>that</nc> are associated with <nc>the same type</nc> of <nc>content</nc>, including (<nc>i</nc>) <nc>a first formula</nc> having <nc>components</nc> based on <nc>the first taxonomy</nc> for calculating <nc>the first common analysis measure</nc> from <nc>electronically tagged financial data</nc> from <nc>the first file</nc>, and <nc>(ii</nc>) <nc>a second formula</nc> having <nc>components</nc> based on <nc>the second taxonomy</nc> for calculating <nc>the first common analysis measure</nc> from <nc>electronically tagged financial data</nc> from <nc>the second file</nc>, wherein <nc>the first common analysis measure</nc> comprises <nc>an analytical metric</nc> not present in <nc>either the first file</nc> or <nc>the second file</nc>; and <nc>the presentation component</nc> comprising: <nc>one or more presentation modules</nc> for presenting <nc>information</nc> associated with <nc>the selected files</nc>, <nc>the presented information</nc> including <nc>data elements</nc> and <nc>a calculation</nc> of <nc>at least the first common analysis measure</nc> automatically formatted according to <nc>presentation information</nc> associated with <nc>the selected files</nc> so as to provide simultaneous line-by-<nc>line</nc> display of <nc>at least the first calculated common analysis measures</nc> <nc>that</nc> correspond to <nc>different taxonomies</nc> and <nc>another calculated common analysis measure</nc>.
| 2
|
2. <nc>The computer program product</nc> of <nc>claim</nc> 1 , wherein <nc>the file access component</nc> includes <nc>a means</nc> for retrieving <nc>the files</nc> with <nc>data elements</nc> corresponding to <nc>a taxonomy</nc> associated with <nc>an industry</nc>, <nc>country</nc>, or <nc>user-created list</nc>.
|
9047261
|
13252463
|
1. <nc>A method</nc> of editing <nc>a document</nc>, <nc>the method</nc> including: a) receiving <nc>data</nc> associated with <nc>the document</nc>, <nc>the data</nc> including <nc>mark-up language data</nc>; b) processing <nc>the received data</nc> to render <nc>at least part</nc> of <nc>the document</nc> for <nc>display</nc> in <nc>a first display area</nc> of <nc>a display</nc>, and displaying as rendered <nc>the at least part</nc> of <nc>the document</nc> in <nc>the first display area</nc>, wherein the rendering comprises formatting <nc>the at least part</nc> of <nc>the document</nc> based on <nc>the mark-up language data</nc>; c) processing <nc>the received data</nc> to render <nc>the at least part</nc> of <nc>the document</nc> for <nc>display</nc> in <nc>a second display area</nc> of <nc>the display</nc>, and displaying as rendered <nc>the at least part</nc> of <nc>the document</nc> in <nc>the second display area</nc>, wherein the rendering comprises formatting <nc>the at least part</nc> of <nc>the document</nc> based on <nc>the mark-up language data</nc>; d) receiving <nc>editing data</nc>; e) editing <nc>the at least part</nc> of <nc>the document</nc> displayed in <nc>the second display area</nc> using <nc>the editing data</nc>, and applying <nc>said editing</nc> to <nc>the at least part</nc> of <nc>the document</nc> displayed in <nc>the first display area</nc>; f) storing <nc>data</nc> associated with <nc>the at least part</nc> of <nc>the document</nc> edited in <nc>e</nc>) as <nc>a first edited version</nc>; g) receiving <nc>further editing data</nc>; h) further editing <nc>the at least part</nc> of <nc>the document</nc> displayed in <nc>the second display area</nc> using <nc>the further editing data</nc> and applying <nc>the further editing</nc> to <nc>the edited at least part</nc> of <nc>the document</nc> displayed in <nc>the first display area</nc>; and <nc>i</nc>) storing as <nc>a second edited version data</nc> associated with <nc>the further edited at least part</nc> of <nc>the document</nc>.
| 24
|
24. <nc>A method</nc> according to <nc>claim</nc> 1 , including accessing <nc>a further document</nc> using <nc>the first display area</nc>.
|
7606428
|
11073690
|
1. <nc>A computer readable recording medium</nc> having encoded <nc>thereon</nc> <nc>an XMT (extensible MPEG-4 textual format</nc>) <nc>schema</nc> for <nc>DIBR data</nc>, <nc>the XMT schema</nc> comprising: <nc>a BitWrapper node schema</nc> used for <nc>graphics data compression</nc>; and <nc>a DepthImage node schema</nc>, <nc>which</nc> is used for <nc>depthimage-based model rendering</nc>, includes <nc>camera information</nc> and <nc>texture information</nc> having <nc>a depth information</nc>, defines <nc>diTexture</nc> as <nc>an element</nc> including <nc>SFDepthTextureNode</nc> as <nc>a model group</nc>, wherein <nc>the BitWrapper node schema</nc> comprises: <nc>a node element</nc>, <nc>which</nc> contains <nc>graphics data</nc> including <nc>data</nc> to be compressed and refers to <nc>SFWorldNode</nc> as <nc>a subelement</nc>; <nc>a BitWrapperEncodingParameter element</nc>; and <nc>three attributes</nc> having <nc>the names</nc> type, <nc>url</nc>, and buffer and <nc>the types</nc> <nc>SFInt32</nc>, <nc>MFUrl</nc>, and <nc>SFString</nc>, respectively, and <nc>the camera information</nc> of <nc>the depthimage node schema</nc> defines at least one of <nc>position</nc>, <nc>orientation</nc>, fieldOfView, <nc>nearPlane</nc>, <nc>farPlane</nc>, and orthographic as <nc>an attribute name</nc>, and attribute <nc>types</nc> defined in <nc>the camera information</nc> include <nc>SFVec3f</nc>, <nc>SFRotation</nc>, <nc>SFVec2f</nc>, <nc>SFFloat</nc>, <nc>SFFloat</nc>, and <nc>SFBool</nc>, respectively.
| 6
|
6. <nc>The computer</nc> readable medium of <nc>claim</nc> 1 , wherein <nc>the XMT schema</nc> further comprises <nc>a decoding information schema</nc> including <nc>an AFXConfig schema</nc> and <nc>a DecoderConfigDescriptor schema</nc> when <nc>the BitWrapper node schema</nc> is <nc>an out-band</nc> using <nc>a url field</nc>, wherein <nc>the DecoderConfigDescriptor schema</nc> defines <nc>an element name</nc> as <nc>AFXconfig</nc> in <nc>a decSpecificlnfo node</nc> in <nc>a DecoderConfigDescriptor node</nc> including <nc>information</nc> required for <nc>decoding</nc>, and <nc>the AFXConfig schema</nc> defines <nc>OctreeImageDecoderSpecific information</nc> related to <nc>use</nc> of <nc>an octree image decoder</nc> for decoding and <nc>PointTextureCompDecoderSpecific information</nc> related to <nc>use</nc> of <nc>a point texture decoder</nc> for decoding as <nc>element names</nc>.
|
6075550
|
08997531
|
1. <nc>A censoring assembly</nc> adapted for <nc>use</nc> with <nc>a signal</nc> having <nc>a closed caption component</nc>, <nc>the censoring assembly comprising</nc>: (<nc>A) closed caption decoding means</nc>, connected to <nc>the signal</nc> having <nc>a closed caption component</nc>, for producing <nc>a video signal</nc> having <nc>no closed caption component</nc> and for converting <nc>the closed caption component</nc> to <nc>an uncensored text data signal</nc>; (<nc>B</nc>) censor device means, connected to <nc>the closed caption decoding means</nc>, for deleting <nc>portions</nc> of <nc>the uncensored text data signal</nc> found to be objectionable, thereby forming <nc>a censored text data signal</nc>; and <nc>(C</nc>) closed caption generator means, connected to <nc>the censored text data signal</nc> and to <nc>the video signal</nc> having <nc>no closed caption component</nc>, for generating <nc>a video signal</nc> having <nc>a censored closed caption component</nc>.
| 2
|
2. <nc>The censoring assembly</nc> of <nc>claim</nc> 1, wherein <nc>the censor device</nc> means further comprises <nc>means</nc> for switching off, for <nc>a timed period</nc>, <nc>an audio signal portion</nc> of <nc>the signal</nc> having <nc>a closed caption component</nc>, in <nc>response</nc> to <nc>the deleting</nc> of <nc>portions</nc> of <nc>the uncensored text data signal</nc> found to be objectionable.
|
8060231
|
12331053
|
1. <nc>One or more computer storage media</nc> including <nc>computer executable instructions</nc> embodied <nc>thereon</nc> <nc>that</nc>, when executed by <nc>one or more computing devices</nc>, cause <nc>the one or more computing devices</nc> to perform <nc>a method</nc> for generating <nc>instructions</nc> for filling <nc>a predetermined region</nc>, <nc>the method</nc> comprising: receiving <nc>a dithered specification</nc> for <nc>the predetermined region</nc>; generating a lattice corresponding to <nc>the dithered specification</nc>, wherein <nc>the lattice</nc> is populated with <nc>a set</nc> of <nc>vertices</nc>, <nc>a set</nc> of <nc>edges</nc> connecting <nc>the vertices</nc>, and cost function data associated with <nc>the edges</nc> on-<nc>the-fly</nc>; recursively performing <nc>an overlapping divide-and-conquer beam search</nc> on <nc>the lattice</nc> to determine <nc>a path</nc> through <nc>a subset</nc> of <nc>vertices</nc> included in <nc>the lattice</nc> <nc>that</nc> minimizes <nc>an overall cost</nc> of filling <nc>the predetermined region</nc>, wherein <nc>the overall cost</nc> is calculated from <nc>the cost function data</nc> associated with <nc>the edges</nc> <nc>that</nc> connect <nc>the subset</nc> of <nc>vertices</nc>; and outputting <nc>instructions</nc> to <nc>a processor</nc> for <nc>the placement</nc> of <nc>building blocks</nc> in <nc>the predetermined region</nc>, <nc>the instructions</nc> generated using <nc>the determined path</nc>.
| 4
|
4. <nc>The one or more computer storage media</nc> of <nc>claim</nc> 1 , wherein <nc>the predetermined region</nc> further comprises <nc>a hollow interior</nc>.
|
9348814
|
14450085
|
1. <nc>A method</nc> for <nc>customer contact handling</nc>, comprising: receiving <nc>contact text</nc> and <nc>contact metadata</nc> from <nc>a user device</nc>; determining <nc>semantic characteristics</nc> of <nc>the contact text</nc> based on <nc>semantic analysis</nc> of <nc>the contact text</nc>; identifying <nc>a user profile</nc> based on <nc>the contact metadata</nc>; retrieving <nc>user data</nc> associated with <nc>the identified user profile</nc>; determining <nc>a contact</nc> need <nc>classification</nc> based on <nc>the semantic characteristics</nc> and <nc>user data</nc>; and selecting <nc>a service agent profile</nc> from <nc>a plurality</nc> of <nc>service agent profiles</nc> based on one or more of <nc>the contact</nc> need <nc>classification</nc>, <nc>semantic characteristics</nc>, and <nc>the data</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein said identifying <nc>the user profile</nc> is further based on <nc>the contact text</nc>.
|
8812540
|
13631663
|
1. <nc>A method</nc> comprising: based on <nc>first content</nc> <nc>that</nc> has been opened within <nc>a content presentation application</nc> executing on <nc>a client device</nc>, <nc>the client device</nc> automatically selecting <nc>context information</nc> to submit to <nc>a server</nc>; responsive to automatically selecting <nc>the context information</nc> for <nc>submission</nc> to <nc>the server</nc>, <nc>the client device</nc> automatically sending <nc>the context information</nc> from <nc>the client device</nc> to <nc>the server</nc>; responsive to sending <nc>the context information</nc> to <nc>the server</nc>, <nc>the client device</nc> receiving <nc>a first search result</nc> from <nc>the server</nc>; responsive to <nc>activation input</nc> activating <nc>a search interface</nc>, displaying <nc>the search interface</nc>; after receiving <nc>the activation input</nc>, and prior to receiving <nc>any user input</nc> of <nc>a query term</nc> via <nc>the activated search interface</nc>, displaying <nc>the first search result</nc> within <nc>a preview section</nc> of <nc>the search interface</nc>; subsequent to displaying <nc>the first search result</nc> in <nc>the preview section</nc>, receiving <nc>user input</nc> entering <nc>one or more query terms</nc> via <nc>the search interface</nc>, <nc>the one or more query terms</nc> including <nc>at least one term</nc> <nc>that</nc> is not found in <nc>the context information</nc>; sending <nc>the one or more query terms</nc> to <nc>the server</nc>; responsive to sending <nc>the one or more query terms</nc> to <nc>the server</nc>, <nc>the client device</nc> receiving <nc>a second search result</nc> from <nc>the server</nc>; displaying <nc>the second search result</nc> in <nc>the search interface</nc> at <nc>the client device</nc>.
| 16
|
16. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the context information</nc> includes one or more of: <nc>link data</nc> identifying <nc>information</nc> about <nc>second content</nc> to <nc>which</nc> the first content links, or category data associated with <nc>the first content</nc>.
|
8180758
|
12118463
|
1. <nc>A computer-implemented method</nc> of providing <nc>a predicate logic corpus</nc>, comprising: under <nc>control</nc> of <nc>one or more computer systems</nc> configured with <nc>executable instructions</nc>, determining <nc>at least one source</nc> of <nc>metadata</nc>; executing <nc>a series</nc> of <nc>Structured Query Language</nc> (<nc>SQL</nc>) queries against <nc>each determined source</nc> of <nc>metadata</nc> to extract <nc>a desired set</nc> of <nc>metadata</nc>; determining <nc>context information</nc> for <nc>the extracted set</nc> of <nc>metadata</nc> using <nc>a predicate logic inference engine</nc>; storing <nc>the extracted set</nc> of <nc>metadata</nc> as <nc>a corpus</nc> of <nc>predicate logic facts</nc> according to <nc>the determined context information</nc>; providing <nc>a user interface</nc> enabling <nc>a user</nc> to define <nc>rules</nc> against <nc>the predicate logic facts</nc> in <nc>the corpus</nc>, and further enabling <nc>the user</nc> to execute <nc>queries</nc> against <nc>the corpus</nc> and <nc>any result</nc> set in <nc>the corpus</nc>, <nc>the queries</nc> being written in at least one of <nc>a natural language</nc> or <nc>a predicate logic notation</nc>, wherein <nc>the predicate logic notation</nc> includes <nc>an artificial intelligence language notation</nc>; receiving <nc>a first query</nc> <nc>that</nc> is created and submitted via <nc>the user interface</nc>; in <nc>response</nc> to receiving <nc>the first query</nc>, executing <nc>the first query</nc> against <nc>the corpus</nc> of <nc>predicate logic facts</nc> obtained through <nc>the determined context information</nc>; and generating <nc>a first result</nc> set to be presented to <nc>the user</nc>, <nc>the generated first result</nc> set including <nc>at least one predicate</nc> defining <nc>a hierarchical relationship</nc> between <nc>data elements</nc> of <nc>the corpus</nc>; and receiving <nc>a second query</nc> <nc>that</nc> is created and submitted via <nc>the user interface</nc>, <nc>the second query</nc> including <nc>at least one new rule</nc> defined by <nc>the user</nc> to be applied against <nc>the generated first result</nc> set via <nc>the user interface</nc>; in <nc>response</nc> to receiving <nc>the second query</nc>, executing <nc>the second query</nc> against <nc>the generated first result set</nc>; and generating <nc>a second result</nc> set to be presented to <nc>the user</nc>, <nc>the generated second result</nc> set being based at least in <nc>part</nc> on <nc>the generated first result set</nc>, wherein when <nc>the generated first result set</nc> includes <nc>at least one set</nc> of <nc>duplicate data elements</nc>, <nc>the generated second result</nc> set excludes <nc>the at least one set</nc> of <nc>duplicate data elements</nc>, <nc>the at least one set</nc> of <nc>duplicate data elements</nc> being excluded based at least in <nc>part</nc> on <nc>the defined hierarchical relationship</nc> between <nc>the data elements</nc> of <nc>the corpus</nc> included in the generated first result set and <nc>the at least one set</nc> of <nc>duplicate data elements</nc>.
| 3
|
3. <nc>The computer-implemented method</nc> according to <nc>claim</nc> 1 , further comprising: inferring <nc>context information</nc> not otherwise stored in <nc>each determined source</nc> of <nc>metadata</nc>.
|
8195085
|
11350266
|
1. <nc>A computerized testing method</nc> comprising <nc>the steps</nc> of<nc>: (a) providing</nc>, in <nc>a computer</nc>, <nc>one or more multiple-choice questions</nc> and displaying <nc>those questions</nc> on <nc>an output device</nc> of <nc>the computer</nc>, each question comprising <nc>a query</nc> and <nc>a plurality</nc> of <nc>answer choices</nc>, wherein said <nc>plurality</nc> of <nc>answer choices</nc> comprises <nc>one correct answer</nc> and <nc>one or more incorrect answers</nc>; (b) for at least one of said <nc>one or more incorrect answers</nc>, prompting <nc>a test-taker</nc> to produce, and enter into <nc>an input device</nc> of <nc>the computer</nc>, <nc>a follow-up query</nc> to <nc>which</nc> said one of said <nc>one or more incorrect answers</nc> is <nc>a correct answer</nc>; (c) determining by <nc>said computer</nc> or by <nc>at least one computer</nc> connected to <nc>the computer</nc> via <nc>a computer network</nc> or <nc>a communication network</nc>, if said one of said <nc>one or more incorrect answers</nc> is <nc>a correct answer</nc> to said <nc>follow-up query</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein said <nc>determining</nc> is by comparing said <nc>follow-up query</nc> with <nc>at least one query</nc> from <nc>a database</nc> of <nc>stored queries</nc> accessible to <nc>the computer</nc> to <nc>which</nc> said one of said <nc>one or more incorrect answers</nc> is <nc>a correct answer</nc>.
|
9116714
|
14256678
|
1. <nc>A method</nc> for <nc>file processing</nc> implemented by <nc>a software compiler</nc> stored in <nc>a non-transitory storage medium</nc>, <nc>the method</nc> comprising: scanning <nc>a source file</nc> and identifying <nc>a target code block</nc> based on <nc>a preset filtering condition</nc>; generating <nc>a first abstract syntax tree</nc> (<nc>AST</nc>) reflecting <nc>a structure</nc> of <nc>the target code</nc> block_and identifying <nc>a plugin position</nc> in <nc>the target code block</nc> using <nc>the first AST</nc>; inserting <nc>a plugin code</nc> into <nc>the first AST</nc> based on <nc>the plugin position</nc>, wherein <nc>the plugin code</nc> is used to monitor <nc>resources</nc> consumed by <nc>the target code block</nc>; generating <nc>a second AST</nc> reflecting <nc>a structure</nc> of <nc>the target code block</nc> after <nc>the plugin code</nc> has been inserted; using <nc>a write-back interface</nc> to write <nc>the second AST</nc> into <nc>the source file</nc>; and invoking <nc>a complier interface</nc> to compile <nc>the source file</nc> with <nc>the plugin codes</nc> produced by <nc>the write-back interface</nc> into <nc>executable files</nc>, wherein executing <nc>the executable files</nc> realizes <nc>functions</nc> of <nc>the target code block</nc> and <nc>functions</nc> of <nc>the plugin code</nc> <nc>such that information</nc> about <nc>the resources</nc> consumed by <nc>the target code block</nc> is obtained; and wherein when receiving <nc>a request</nc> from <nc>a user</nc>, <nc>the source file</nc> with <nc>the plugin codes</nc> produced by <nc>the write-back interface</nc> is further modified by <nc>the user</nc>.
| 2
|
2. <nc>The method</nc> according to <nc>claim</nc> 1 , wherein identifying <nc>the plugin position</nc> in <nc>the target code</nc> <nc>further comprises</nc>: analyzing <nc>syntax</nc> of <nc>the target code block</nc> and generating <nc>the first AST</nc> reflecting <nc>the structure</nc> of <nc>the target code block</nc> based on <nc>the analysis</nc>; scanning <nc>the first AST</nc> and identifying <nc>a plugin position</nc> on <nc>the first AST</nc> <nc>that</nc> corresponds to <nc>a function entry and exit position</nc> in <nc>the target code block</nc>; and identifying <nc>the plugin position</nc> in <nc>the target code block</nc> based on <nc>the plugin position</nc> identified on <nc>the first AST</nc>.
|
8977971
|
13043007
|
1. <nc>A computer-implemented method</nc> of automatically generating <nc>metadata</nc> used to develop <nc>graphical user interfaces</nc>, comprising: scanning <nc>a hand-drawn image</nc> of <nc>a user interface</nc> to detect <nc>a candidate region</nc> on <nc>the image</nc>, <nc>the candidate region</nc> being potentially associated with <nc>a user interface component</nc>; analyzing <nc>the candidate region</nc> to identify <nc>a user interface component</nc> contained therein by comparing <nc>a potential user interface component</nc> to <nc>user interface definitions</nc> stored in <nc>a database</nc> to identify one of <nc>the user interface definitions</nc> associated therewith; extracting <nc>one or more properties</nc> of <nc>the identified user interface component</nc>; generating <nc>metadata</nc> based on <nc>the user interface definition</nc> associated with <nc>the identified user interface component</nc> and <nc>the extracted properties</nc>; and <nc>computer</nc> generating <nc>a second user interface</nc> based on <nc>the metadata</nc> and <nc>the scanned image</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein one or more of <nc>the extracted properties</nc> comprise <nc>static properties</nc>.
|
9378239
|
14858208
|
1. <nc>A computer-system-implemented method</nc> for requesting <nc>desired information</nc> from <nc>a graph database</nc>, <nc>the method</nc> comprising: executing <nc>a query</nc> against <nc>the graph database</nc> storing <nc>a graph</nc>, wherein: <nc>the graph</nc> comprises <nc>nodes</nc>, edges between <nc>the nodes</nc>, and predicates to represent and store <nc>data</nc> with <nc>index-free adjacency</nc>; and <nc>the query</nc> identifies <nc>an edge</nc> associated with <nc>a predicate</nc> <nc>that</nc> specifies one or more of <nc>the nodes</nc> in <nc>the graph</nc>; receiving <nc>a result</nc> in <nc>response</nc> to <nc>the query</nc>, wherein <nc>the result</nc> includes <nc>a subset</nc> of <nc>the graph</nc>; and verifying <nc>the subset</nc> of <nc>the graph</nc>, wherein <nc>the verifying</nc> comprises: executing <nc>another query</nc> against <nc>the graph database</nc>; receiving <nc>another result</nc> in <nc>response</nc> to <nc>the query</nc>, wherein <nc>the other result</nc> includes <nc>another subset</nc> of <nc>the graph</nc>; and comparing <nc>the other subset</nc> of <nc>the graph</nc> with <nc>the subset</nc> of <nc>the graph</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the subset</nc> of <nc>the graph</nc> is verified when there is <nc>a match</nc> between <nc>an overlapping portion</nc> of <nc>the subset</nc> of <nc>the graph</nc> and <nc>the other subset</nc> of <nc>the graph</nc>.
|
10095471
|
14475344
|
1. <nc>A method</nc> for facilitating <nc>user interaction</nc> with <nc>enterprise software</nc> via <nc>a mobile computing device</nc>, <nc>the mobile computing device</nc> including <nc>a runtime memory</nc>, <nc>the mobile computing device</nc> further coupled to <nc>a server memory</nc> via <nc>a network connection</nc>, <nc>the method</nc> comprising: receiving <nc>language input</nc> responsive to <nc>one or more prompts</nc>; determining, based on <nc>the language input</nc>, <nc>a subject category</nc>: determining that <nc>an opportunity object</nc> including <nc>enterprise information</nc> is associated with <nc>the subject category</nc>; checking whether <nc>the opportunity object</nc> is in <nc>the runtime memory</nc> and, if not, then retrieving <nc>the opportunity object</nc> from <nc>the server memory</nc> and storing <nc>it</nc> in <nc>the runtime memory</nc>; maintaining <nc>the opportunity object</nc> in <nc>the runtime memory</nc> and using <nc>the enterprise information</nc> in one or more of <nc>the following acts</nc>; obtaining <nc>subject context information</nc>; identifying <nc>an action category</nc>; employing <nc>identification</nc> of <nc>the action category</nc> to obtain <nc>action context information</nc> pertaining to <nc>the action category</nc>; and implementing <nc>a software action</nc> in <nc>accordance</nc> with <nc>the action context information</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the method</nc> further includes receiving as <nc>input, identity information</nc> identifying <nc>a user</nc> of <nc>software</nc> used to implement <nc>the method</nc>.
|
7536631
|
10452229
|
1. <nc>An apparatus</nc> for providing <nc>verified communication</nc>, comprising: <nc>a transmitter</nc> including: <nc>input terminals</nc> to receive <nc>a data word</nc>, <nc>an encoder</nc> configured to encode <nc>the data word</nc> to create <nc>an encoded word</nc> different from <nc>the data word</nc>, and <nc>output terminals</nc> configured to transmit <nc>the data word</nc> and <nc>the encoded word</nc>; and, <nc>a receiver</nc> coupled to <nc>the transmitter</nc> and including: <nc>input terminals</nc> to receive <nc>the data word</nc> as <nc>a received word</nc> and <nc>the encoded word</nc>, <nc>a decoder</nc> configured to decode <nc>the encoded word</nc> to create <nc>a decoded word</nc>, and <nc>a comparator</nc> configured to compare <nc>the received word</nc> and <nc>the decoded word</nc> to create <nc>a select signal</nc>, and <nc>a selector</nc> responsive to <nc>the select signal</nc> and configured to select <nc>the received data word</nc> or <nc>the decoded word</nc> based at least in <nc>part</nc> on <nc>the select signal</nc>; wherein <nc>the transmitter</nc> is configured to simultaneously transmit <nc>the data word</nc> and <nc>the encoded word</nc>; and wherein <nc>the receiver</nc> is configured to simultaneously receive <nc>the data word</nc> and <nc>the encoded word</nc>; <nc>a first signal detector</nc> coupled to <nc>a first communications channel</nc> including <nc>a set</nc> of <nc>received words</nc>; <nc>a second signal detector</nc> coupled to <nc>a second communications channel</nc> including <nc>a set</nc> of <nc>the encoded words</nc>; wherein, if <nc>the first communications channel</nc> is not available, <nc>the first signal detector</nc> notifies <nc>the comparator</nc> <nc>which</nc> instructs <nc>the selector</nc> to choose <nc>the set</nc> of <nc>encoded words</nc> from <nc>the second communications channel</nc>; and, wherein, if <nc>the second communications channel</nc> is not available, <nc>the second signal detector</nc> notifies <nc>the comparator</nc> <nc>which</nc> instructs <nc>the selector</nc> to choose <nc>the set</nc> of <nc>received words</nc> from <nc>the first communications channel</nc>.
| 7
|
7. <nc>The apparatus</nc> of <nc>claim</nc> 1 , wherein: <nc>the comparator</nc> is configured to compare <nc>the bits</nc> in <nc>the received word</nc> with <nc>those</nc> in <nc>the decoded word</nc> and to select <nc>the valid bits</nc> in <nc>each word</nc> to recreate <nc>the data word</nc>.
|
8131768
|
12331344
|
1. <nc>A computer</nc> implemented <nc>method</nc> for analyzing <nc>data-flow</nc> using <nc>program expressions</nc> or <nc>terms</nc>, comprising: <nc>a.</nc> extracting <nc>a control flow graph node</nc> from <nc>a work list</nc>; b. representing <nc>a symbolic state</nc> as <nc>a condition-map pair</nc> <nc>(C</nc>, σ) where <nc>C</nc> denotes <nc>a current path condition predicate</nc> and <nc>σ</nc> denotes <nc>a map</nc> from <nc>program variables</nc> to <nc>symbolic values</nc>; <nc>c.</nc> merging <nc>symbolic term values</nc> at <nc>join nodes</nc> to avoid <nc>path-explosion</nc> by using choose and <nc>if-then-else (ite) function operators</nc>, where choose is <nc>a term</nc> of <nc>form</nc> choose ((<nc>C</nc> 1 ,t 1 ),(C 2 ,t 2 ),(C 3 ,t 3 )) on <nc>expression sort</nc> with <nc>a non-deterministic choice</nc> between <nc>the values</nc> <nc>t</nc> 1 (1≦i≦3) given <nc>a corresponding condition C</nc> 1 ; <nc>d.</nc> performing <nc>simplification</nc> of <nc>term values</nc> using <nc>rewriting logic</nc>, wherein <nc>a set</nc> of <nc>rules</nc> for simplifying <nc>choose</nc> and <nc>ite terms</nc> are used along with <nc>rules</nc> for simplifying <nc>Presburger arithmetic expressions</nc>; <nc>e.</nc> determining <nc>successors</nc> of <nc>the graph node</nc> to <nc>which</nc> <nc>data</nc> must be propagated; <nc>f.</nc> updating <nc>symbolic data</nc> for <nc>elements</nc> of <nc>the successors</nc>; <nc>g.</nc> performing anti<nc>-</nc><nc>unification</nc> to generalize <nc>similar terms</nc> obtained at <nc>a loop head</nc>; and h. displaying <nc>the program analysis</nc> for <nc>code review</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the successors</nc> are determined using <nc>a Satisfiability Modulo Theories (SMT) solver</nc>.
|
9524717
|
14514918
|
1. <nc>An interface</nc> for <nc>a call system</nc>, <nc>the interface</nc> comprising: <nc>a telephony engine</nc> configured to: receive <nc>a voice call</nc> from <nc>the call system</nc>, <nc>the voice call</nc> including <nc>a spoken request</nc>; automatically answer <nc>the voice call</nc>: if answering <nc>the voice call</nc> would not increase <nc>the number</nc> of <nc>concurrently answered voice</nc> calls to more than <nc>a pre-determined maximum number</nc> of <nc>concurrently answered voice calls</nc>, and only if <nc>the voice call</nc> is not answered after <nc>a pre-determined number</nc> of <nc>rings</nc>; transfer <nc>the voice call</nc> to <nc>a pre-determined call center</nc> if answering <nc>the voice call</nc> would increase <nc>the maximum number</nc> of <nc>concurrently answered voice</nc> calls to more than <nc>the pre-determined maximum number</nc> of <nc>concurrently answered voice calls</nc>; a voice recognition engine coupled with <nc>the telephony engine</nc> for analyzing <nc>the voice call</nc> and generating <nc>text data representative</nc> of <nc>at least a portion</nc> of <nc>the request</nc>, <nc>the call system</nc> being configured to: transmit <nc>the text data</nc> to <nc>a mobile device</nc> of <nc>a nurse</nc> or <nc>caregiver</nc> for responding to <nc>the patient request</nc> if converting <nc>the at least a portion</nc> of <nc>the patient request</nc> into <nc>text data</nc> is successful, and transmit <nc>the patient request</nc> to <nc>the mobile device</nc> of <nc>the nurse</nc> or <nc>caregiver</nc> in <nc>the form</nc> of <nc>a recording</nc> of <nc>the voice call</nc> if converting <nc>the at least a portion</nc> of <nc>the patient request</nc> into <nc>text data</nc> is unsuccessful; and <nc>an analytics and reporting engine</nc> configured to analyze <nc>a success rate</nc> of voice-to-<nc>text</nc> conversion performed by <nc>the voice recognition engine</nc> and report <nc>results</nc> of <nc>the analysis</nc>.
| 2
|
2. <nc>The interface</nc> of <nc>claim</nc> 1 , further comprising <nc>an alarm and routing engine</nc> for formatting <nc>the text data</nc> and for directing <nc>the text data</nc> to <nc>an alert management system</nc> configured to route <nc>the text data</nc> to <nc>a mobile device</nc> or call <nc>center</nc> for responding to <nc>the request</nc>, wherein <nc>the alarm and routing engine</nc> is configured to determine whether <nc>a recording</nc> of <nc>the voice call</nc> will be routed to <nc>an intended recipient</nc> or <nc>the voice call</nc> will be routed to <nc>a predetermined call center</nc> for interpreting <nc>the request</nc> if <nc>the voice recognition engine</nc> is unable to generate <nc>text data representative</nc> of <nc>at least a portion</nc> of <nc>the request</nc>.
|
9710463
|
14099079
|
1. <nc>A computer-implemented method</nc> for <nc>linguistic processing</nc> for speech-to-<nc>speech</nc> translation, <nc>the method</nc> comprising: receiving <nc>a linguistic input</nc> comprising <nc>a sequence</nc> of <nc>words</nc> in <nc>a first language</nc> from <nc>a first user</nc>, <nc>the linguistic input</nc> comprising <nc>a first audio input</nc> including <nc>a speech utterance</nc> by <nc>the first user</nc>; determining <nc>a first data representation</nc> of <nc>the linguistic input</nc>; processing, using <nc>a computer-implemented analyzer</nc>, <nc>the first data representation</nc> to identify <nc>at least part</nc> of <nc>the data representation</nc> as being potentially associated with <nc>an error</nc> of <nc>processing</nc> of <nc>the linguistic input</nc>, wherein <nc>the processing</nc> comprises identifying said <nc>part</nc> as <nc>at least one characteristic</nc> of <nc>(a</nc>) including out-of-<nc>vocabulary</nc> (OOV) words, (b) representing <nc>a named entity</nc>, (c) including <nc>a homophone</nc>, (d) having <nc>an ambiguous word sense</nc>, and (e) including <nc>an idiom</nc> in <nc>the first language</nc>; performing <nc>further processing</nc>, using <nc>a computer-implemented recovery strategy processor</nc>, of <nc>the identified at least part</nc> of <nc>the first data representation</nc> to form <nc>a modified data representation</nc> of <nc>the linguistic input</nc>; using <nc>a machine translator</nc> to form <nc>a second data representation</nc> of <nc>the modified data representation</nc>; and processing <nc>the second data representation</nc> using <nc>the recovery strategy processor</nc> to refine <nc>the second data representation</nc> through at least one of <nc>automated processing</nc> or <nc>user-assisted processing</nc>; determining <nc>a linguistic output</nc> from <nc>the refined second data representation</nc>, <nc>the linguistic output</nc> comprising <nc>a sequence</nc> of <nc>words</nc> in <nc>a second language</nc>; and providing <nc>the linguistic output</nc> to <nc>a second user</nc>, <nc>the linguistic output</nc> comprising <nc>a synthesized second audio signal</nc> including <nc>speech output</nc>, wherein identifying said part as including <nc>an idiom</nc> in <nc>the first language</nc> comprises performing <nc>rule-based idiom expansion</nc> and performing <nc>statistical idiom detection</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 wherein <nc>the first data representation</nc> comprises <nc>a text representation</nc> in <nc>the first language</nc>.
|
8914718
|
12773402
|
1. <nc>Method</nc> of coding <nc>a structured document</nc> comprising <nc>events</nc> to be coded having <nc>values</nc>, comprising <nc>the following steps</nc>: running through <nc>the document</nc> in <nc>order</nc> to process <nc>the document event</nc> by <nc>event</nc>; forming, in <nc>memory</nc>, <nc>channels</nc> of <nc>values</nc> containing <nc>values</nc> of <nc>events</nc> according to <nc>at least one criterion</nc>; once <nc>the channels</nc> of <nc>values</nc> have been formed, coding <nc>the channels</nc> of <nc>values</nc> thus formed by coding <nc>the event values</nc> of <nc>each</nc> of <nc>these channels</nc> of <nc>values</nc> by <nc>means</nc> of <nc>at least one coding dictionary</nc>; wherein <nc>the formation step</nc> comprises, for <nc>each event</nc> being processed and having <nc>a value</nc>, identifying <nc>an entry</nc> of <nc>the coding dictionary</nc> <nc>that</nc> associates <nc>the value</nc> of <nc>the event</nc> with <nc>a coding index</nc>, and adding <nc>this value</nc> to one of <nc>said channels</nc> of <nc>values</nc> by storing, in <nc>memory</nc> storing <nc>the channel</nc> of <nc>values</nc>, <nc>a reference</nc> to <nc>the identified entry</nc> instead of storing <nc>the event value</nc> <nc>itself</nc>, and wherein <nc>the coding</nc> of <nc>the channels</nc> includes running though each reference stored in <nc>each channel</nc> of <nc>values</nc>, using <nc>the reference</nc> run through to retrieve <nc>a coding index</nc> or <nc>a coding value</nc> from <nc>the corresponding entry</nc> of <nc>the coding dictionary</nc>, and adding <nc>the retrieved coding index</nc> or <nc>value</nc> to <nc>a coded bitstream</nc>.
| 2
|
2. <nc>Method</nc> according to <nc>claim</nc> 1 , wherein at least one of <nc>the addings</nc> performed if <nc>a value</nc> to be added is not already there.
|
6161084
|
09366499
|
1. <nc>A method</nc> in <nc>a computer system</nc> for generating <nc>information retrieval tokens</nc> from <nc>an input string</nc>, <nc>the method</nc> comprising <nc>the steps</nc> of: creating from <nc>the input string</nc> <nc>a primary logical form</nc> characterizing <nc>a semantic relationship</nc> between <nc>selected words</nc> in <nc>the input string</nc>; identifying <nc>hypernyms</nc> of <nc>the selected words</nc> in <nc>the input string</nc>, <nc>at least one hypernym</nc> being identified from <nc>a group</nc> of <nc>hypernyms</nc> associated with <nc>a selected word</nc> wherein <nc>at least one other hypernym</nc> in <nc>the group</nc> is not identified; constructing from <nc>the primary logical form</nc> <nc>one or more alternative logical forms</nc>, <nc>each alternative logical form</nc> being constructed by, for <nc>each</nc> of one or more of <nc>the selected words</nc> in <nc>the input string</nc>, replacing <nc>the selected word</nc> in <nc>the primary logical form</nc> with <nc>an identified hypernym</nc> of <nc>the selected word</nc>; and generating <nc>tokens</nc> representing <nc>both the primary logical form</nc> and <nc>the alternative logical forms</nc>, <nc>the generated tokens</nc> being distinguishable by <nc>an information retrieval engine</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1, further comprising <nc>the steps</nc> of: before <nc>the constructing step</nc>, selecting <nc>the input string</nc> from <nc>a body</nc> of <nc>text</nc> to be indexed; and submitting <nc>the generated tokens</nc> to <nc>an indexing subsystem</nc> for <nc>storage</nc> in <nc>an index</nc> representing <nc>the body</nc> of <nc>text</nc>.
|
9569327
|
13644187
|
1. <nc>An alert processing system</nc> comprising: <nc>a communication interlace</nc> adapted to receive <nc>at least one alert description</nc> in <nc>a natural language</nc> of <nc>an alert message output</nc> by <nc>an associated device</nc>, <nc>the alert description</nc> defining <nc>a problem</nc> with <nc>the associated device</nc>; <nc>an alert filtering module</nc> adapted to: access <nc>a stored referenced alerts list</nc> to determine if <nc>the alert description</nc> corresponds to one of <nc>a set</nc> of <nc>previously referenced alert descriptions</nc>, when <nc>the alert description</nc> is determined to correspond to <nc>a previously referenced alert description</nc>, labeling <nc>the alert description</nc> with <nc>the label</nc> of <nc>the corresponding previously referenced alert description</nc>, and when <nc>the alert description</nc> is determined not to correspond to <nc>any</nc> of <nc>the previously referenced alert descriptions</nc>, labeling <nc>the alert description</nc> as <nc>an unreferenced alert description</nc>; <nc>a word exclusion module</nc> adapted to extract <nc>a set</nc> of <nc>words</nc> from <nc>the alert description</nc> related to <nc>a condition</nc> of <nc>the associated device</nc>; <nc>an alert categorization module</nc> adapted to categorize <nc>the alert description</nc>, when <nc>it</nc> has been labeled as unreferenced by <nc>the alert filtering module</nc>, into to one of <nc>a predetermined set</nc> of <nc>alert categories</nc> of <nc>a categorization model</nc> based on <nc>the extracted set</nc> of <nc>words</nc>; <nc>a language module</nc> adapted to: compare <nc>each word</nc> in <nc>the set</nc> of <nc>words</nc> with <nc>a plurality</nc> of <nc>stored dictionaries</nc> in <nc>different languages</nc> for identifying <nc>at least one source language</nc> of <nc>the word</nc>, assign <nc>a source language</nc> to <nc>the set</nc> of <nc>words</nc> based on <nc>the identified source languages</nc> of <nc>the words</nc>, and translate <nc>the set</nc> of <nc>words</nc> from <nc>the assigned source language</nc> to <nc>a target language</nc>; and <nc>a processor</nc> adapted to implement <nc>the modules</nc>.
| 5
|
5. <nc>The alert processing system</nc> of <nc>claim</nc> 1 , wherein <nc>the categorization module</nc> is trained to categorize <nc>sets</nc> of <nc>words</nc> in <nc>the target language</nc>.
|
9275291
|
13919937
|
1. <nc>A system</nc> for <nc>enhanced machine learning</nc> using <nc>coder rankings</nc>, <nc>the system</nc> comprising: <nc>a first physical memory device</nc> configured to store <nc>a plurality</nc> of <nc>data sets</nc> including <nc>a first data</nc> set corresponding to <nc>a first subject area category</nc>, <nc>each</nc> of <nc>the plurality</nc> of <nc>data sets</nc> including at least one training data set; <nc>a second physical memory device</nc> configured to store <nc>coder data</nc> related to <nc>users</nc> <nc>who</nc> perform <nc>data coding</nc>; <nc>a classification computer processor module</nc> in <nc>electronic communication</nc> with <nc>the first and second physical memory device</nc>, <nc>the classification computer processor module</nc> configured to: store <nc>a first set</nc> of <nc>selected codes</nc> corresponding to <nc>a set</nc> of <nc>data items</nc> in <nc>a training data</nc> set of the first data set, <nc>the first set</nc> of <nc>selected codes</nc> provided by <nc>a first user</nc> via <nc>a first classifier system</nc>; store <nc>a second set</nc> of <nc>selected codes</nc> corresponding to <nc>the set</nc> of <nc>data items</nc>, the second set of <nc>selected codes</nc> provided by <nc>a second user</nc> via <nc>a second classifier system</nc>, wherein <nc>the first set</nc> of <nc>selected codes</nc> is different from <nc>the second set</nc> of <nc>selected codes</nc>; and electronically access <nc>an adjudicated set</nc> of <nc>selected codes</nc> adjudicated by <nc>a third user</nc> via <nc>an adjudication system</nc>; and a coder ranking processor module in <nc>electronic communication</nc> with <nc>the first and second physical memory device</nc>, <nc>the coder</nc> ranking processor module configured to: generate <nc>a first trust score</nc> corresponding to <nc>the first user</nc> as to <nc>the first subject area category</nc>, <nc>the first trust score</nc> based at least in <nc>part</nc> on: <nc>an accuracy determination</nc> of <nc>the first set</nc> of <nc>selected codes</nc> as compared to <nc>the adjudicated set</nc> of <nc>selected codes</nc>; and at least two or more of <nc>the following</nc>: <nc>an accuracy determination</nc> of <nc>previous coding performance</nc> by <nc>the first user</nc> from <nc>a prior time period</nc>; <nc>third party accreditation</nc> for <nc>the first user</nc> related to <nc>the first subject area category</nc>, or <nc>third party credentials</nc> for <nc>the first user</nc> related to <nc>the first subject area category</nc>; generate <nc>a second trust score</nc> corresponding to <nc>the second user</nc> as to <nc>the first subject area category</nc>, <nc>the second trust score</nc> based at least in <nc>part</nc> on: <nc>an accuracy determination</nc> of <nc>the second set</nc> of <nc>selected codes</nc> as compared to <nc>the adjudicated set</nc> of <nc>selected codes</nc>; and at least two or more of <nc>the following</nc>: <nc>an accuracy determination</nc> of <nc>previous coding performance</nc> by <nc>the second user</nc> from <nc>a prior time period</nc>; <nc>third party accreditation</nc> for <nc>the second user</nc> related to <nc>the first subject area category</nc>, or <nc>third party credentials</nc> for <nc>the second user</nc> related to <nc>the first subject area category</nc>; wherein <nc>the accuracy determination</nc> of <nc>the first set</nc> of <nc>selected codes</nc> indicates <nc>a level</nc> of <nc>accuracy</nc> higher than <nc>the accuracy determination</nc> of <nc>the second set</nc> of <nc>selected codes</nc> and <nc>the first trust score</nc> corresponding to <nc>the first user</nc> is higher than <nc>the second trust score</nc> corresponding to <nc>the second user</nc>; and store <nc>the first trust score</nc> and <nc>the second trust score</nc> in <nc>the second physical memory device</nc>.
| 11
|
11. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the first user</nc> is the same as <nc>the third user</nc>.
|
8210851
|
11464590
|
1. In <nc>a computer-implemented auditory training exercise</nc> for improving <nc>a person's acoustic processing abilities</nc>, wherein <nc>the auditory training exercise</nc> adaptively modifies <nc>a stimulus level</nc> of <nc>synthesized phonemes</nc> in <nc>an iterative manner</nc> based on <nc>the person's responses</nc> to <nc>those synthesized phonemes</nc>, wherein <nc>the stimulus level</nc> is characterized by one or more of <nc>the degree</nc> by <nc>which</nc> selected <nc>portions</nc> of <nc>the synthesized phonemes</nc> are made relatively more or less loud and <nc>the degree</nc> by <nc>which</nc> <nc>selected segments</nc> of <nc>the synthesized phonemes</nc> are temporally stretched, <nc>a method</nc> of progressively orienting <nc>the person</nc> to recognize <nc>a synthesized phoneme</nc> <nc>that</nc> consists essentially only of <nc>a formant-filtered representation</nc> of <nc>the phoneme</nc>, <nc>the method</nc> comprising: providing <nc>a first, relatively simplified representation</nc> of <nc>a phoneme</nc>, <nc>the first representation</nc> consisting of <nc>a formant-synthesized-filtered representation</nc> of <nc>the phoneme</nc>; providing <nc>a second relatively more spectrally complete representation</nc> of <nc>the phoneme</nc> thru <nc>synthesis</nc> using <nc>a naturalistic time-varying filter</nc>; blending <nc>different proportions</nc> of <nc>the first and second representations</nc> of <nc>the phoneme</nc> into <nc>different blended representations</nc> of <nc>the phoneme</nc> ranging from <nc>a most naturally-sounding representation</nc> of <nc>the phoneme</nc> to <nc>a least naturally-sounding representation</nc> of <nc>the phoneme</nc>; initially aurally presenting <nc>the most naturally-sounding representation</nc> of <nc>the phoneme</nc> to <nc>the person</nc>; and aurally presenting <nc>progressively less-naturally sounding representations</nc> of <nc>the phoneme</nc> to <nc>the person</nc>; wherein <nc>the auditory training exercise</nc> is implemented after <nc>the person</nc> completes <nc>the orientation</nc>.
| 23
|
23. <nc>The method</nc> as recited in <nc>claim</nc> 1 , <nc>wherein the computer-implemented auditory training exercise</nc>, after detecting <nc>an incorrect response</nc>, presents <nc>a synthesized phoneme</nc> with <nc>a stimulus level</nc> <nc>that</nc> has <nc>greater stretching</nc>.
|
7774370
|
11763209
|
1. <nc>A method</nc> for managing <nc>a message communication system</nc> comprising: providing <nc>a token log</nc> comprising <nc>a data structure</nc> <nc>that</nc> contains <nc>a user-extensible set</nc> of <nc>tokens</nc> and <nc>implied conditions</nc> under <nc>which</nc> <nc>specific tokens</nc> are considered valid; creating <nc>a new token</nc> comprising <nc>a set</nc> of <nc>symbols</nc> provided by <nc>a system operator</nc> <nc>that</nc> together have <nc>no prior functional meaning</nc> within <nc>the message communication system</nc> outside of <nc>the token log</nc>; storing <nc>the new token</nc> in <nc>the token log</nc> together with <nc>information</nc> <nc>that</nc> implies <nc>one or more conditions</nc> under <nc>which</nc> <nc>the token</nc> is considered valid, which <nc>validity information</nc> implies <nc>one or more specific actions</nc> to be performed when a message accompanied by <nc>the new token experiences</nc> an event; providing for distributing <nc>the new token</nc> to <nc>one or more external entities</nc>; and providing <nc>a system operator</nc> with <nc>the ability</nc> to change <nc>a validity status</nc> of <nc>the token</nc>; receiving <nc>an incoming message</nc> with <nc>the new token</nc> attached; determining <nc>the validity status</nc> of <nc>the new token</nc> by attempting to locate <nc>the token</nc> in <nc>the token log</nc> and checking <nc>the one or more conditions</nc> for <nc>use</nc>; and rejecting <nc>the message</nc> if <nc>the new token</nc> was determined to no longer be valid.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein providing for distributing <nc>the new token</nc> to <nc>one or more external entities</nc> comprises including <nc>the new token</nc> with <nc>one or more outgoing messages</nc> <nc>that</nc> are sent to <nc>the one or more external entities</nc>.
|
9886958
|
14966257
|
1. <nc>A computing device</nc> comprising: <nc>at least one processor</nc>; and <nc>a memory</nc> for storing and <nc>encoding computer</nc> <nc>executable instructions</nc> <nc>that</nc>, when executed by <nc>the at least one processor</nc> are operative to: receive <nc>an utterance</nc> from <nc>a user</nc>; extract <nc>features</nc> from <nc>the utterance</nc> to form <nc>extracted features</nc>; identify that <nc>the utterance</nc> is directed to at least one of <nc>a plurality</nc> of <nc>items</nc> previously provided by <nc>the computing device</nc> to <nc>the user</nc> utilizing <nc>a language independent disambiguation model</nc>, wherein <nc>the language independent disambiguation model</nc> identifies that <nc>the utterance</nc> is directed to the at least one of <nc>the plurality</nc> of <nc>items</nc> by: identifying as universal <nc>features</nc> <nc>one or more domain independent features</nc> and <nc>language independent features</nc> in <nc>the extracted features</nc>; determining <nc>an overlap</nc> between <nc>one or more universal features</nc> extracted from <nc>the utterance</nc> and <nc>one or more features</nc> associated with <nc>the plurality</nc> of <nc>items</nc>, wherein <nc>the one or more features</nc> are identified based on <nc>metadata</nc> associated with <nc>the plurality</nc> of <nc>items</nc>; and identifying the at least one of <nc>the plurality</nc> of <nc>items</nc> corresponding to <nc>the utterance</nc> based on <nc>the overlap</nc>; and send <nc>instructions</nc> to perform <nc>an action</nc> associated with <nc>the utterance</nc> upon identifying that <nc>the utterance</nc> is directed to the at least one of <nc>the plurality</nc> of <nc>items</nc>.
| 2
|
2. <nc>The computing device</nc> of <nc>claim</nc> 1 , wherein <nc>the utterance</nc> is in <nc>a first language</nc>, and wherein <nc>the language independent disambiguation model</nc> is trained utilizing <nc>data</nc> from <nc>a different language</nc> from <nc>the first language</nc>.
|
8631386
|
11211043
|
1. <nc>A system</nc> for developing <nc>web services</nc>, comprising: <nc>a Design Time Framework</nc> including <nc>a code generator</nc> <nc>that</nc> receives <nc>a first set</nc> of <nc>configuration files</nc> from <nc>a user</nc> and generates <nc>source code artifacts</nc> in <nc>an object oriented programming language</nc> based on <nc>the received first set</nc> of <nc>configuration files</nc>, wherein <nc>the first set</nc> of <nc>configuration files</nc> comprises <nc>an XML schema file</nc>, <nc>an XML descriptor file</nc>, and <nc>an XML rule file</nc>, wherein <nc>the XML schema file</nc> comprises <nc>XML schema annotations</nc> specifying <nc>properties</nc> of <nc>fields</nc> in <nc>the XML schema file</nc>, <nc>wherein an output</nc> of <nc>the Design Time Framework results</nc> from <nc>a compiling</nc> of <nc>the source code artifacts</nc> in <nc>an object oriented programming language</nc> based on <nc>the XML schema file</nc>, and <nc>the output</nc> comprises <nc>binary objects</nc> in <nc>a binary format</nc> and having <nc>field properties</nc> specified by <nc>the XML schema annotations</nc>; <nc>an Object-Service Framework</nc> <nc>that</nc> receives <nc>a second set</nc> of <nc>configuration files</nc> from <nc>the user</nc>, wherein <nc>the Object-Service Framework</nc> comprises <nc>a set</nc> of <nc>pre-built runtime services</nc>, wherein <nc>each</nc> of <nc>the pre-built runtime services</nc> extends <nc>the Object-Service Framework</nc> and comprises <nc>binary code</nc> <nc>that</nc> was compiled before <nc>the compiling</nc> of <nc>the source code artifacts</nc> of <nc>the Design Time Framework</nc> in <nc>an object oriented programming language</nc> based on <nc>the XML schema file</nc>, wherein <nc>behaviors</nc> of <nc>the pre-built runtime services</nc> can be dynamically changed using <nc>the second set</nc> of <nc>configuration files</nc>, and wherein <nc>the Object Service Framework</nc> defines <nc>an application program interface</nc> (<nc>API</nc>) for <nc>each</nc> of <nc>the pre-built runtime services</nc> and wherein <nc>the pre-built runtime services</nc> interact with <nc>the source code artifacts</nc> of <nc>the Design Time Framework</nc> through <nc>the application program interface</nc> and wherein <nc>the pre-built runtime services</nc> extend <nc>the application program interface</nc>, and for <nc>CRUD services</nc> of <nc>the pre-built runtime services</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> to perform on <nc>a persistent database</nc> of <nc>the system</nc> at least one of <nc>create</nc>, retrieve, <nc>update</nc>, or delete <nc>instances</nc> of <nc>information structures</nc>, wherein <nc>information</nc> is passed between <nc>the CRUD services</nc> and <nc>source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>, for <nc>security field services</nc> of <nc>the pre-built runtime services</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> and obtains <nc>security policy types</nc>, <nc>security policy operations</nc>, and <nc>resource values</nc> in <nc>the second set</nc> of <nc>configuration files</nc> to control <nc>the access</nc> to <nc>the web service</nc> and <nc>fields</nc> of <nc>the web service</nc>, wherein <nc>information</nc> is passed between <nc>the security field services</nc> and <nc>source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>, for <nc>an XML marshall service</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> and retrieves <nc>dynamic properties</nc> specified in <nc>the second set</nc> of <nc>configuration files</nc> to create <nc>XML formatted data</nc> having <nc>a first XML format</nc>, wherein <nc>information</nc> is passed between <nc>the XML marshall service and source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>, for <nc>an XML transformation service</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> to transform <nc>the created XML formatted data</nc> having <nc>the first XML format</nc> to <nc>XML formatted data</nc> having <nc>a second XML format</nc>, different from <nc>the first XML format</nc>, wherein <nc>information</nc> is passed between <nc>the XML transformation service</nc> and <nc>source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>, for <nc>an XML schema validation service</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> and retrieves <nc>dynamic properties</nc> specified in <nc>the second set</nc> of <nc>configuration files</nc> to validate and convert <nc>an input XML formatted data</nc> into <nc>an instance</nc> of <nc>the binary form</nc> of <nc>the Design Time Framework</nc>, wherein <nc>information</nc> is passed between <nc>the XML schema validation service</nc> and <nc>source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>, and for <nc>a key value service</nc> <nc>that</nc> extend <nc>the Object-Service Framework</nc>, <nc>the Object-Service Framework</nc> searches <nc>the output</nc> of <nc>the Design Time Framework</nc> and retrieves <nc>dynamic properties</nc> specified in <nc>the second set</nc> of <nc>configuration files</nc> to validate and convert <nc>data</nc> formatted as <nc>name-value pair</nc> into <nc>an instance</nc> of <nc>the binary output form</nc> of <nc>the Design Time Framework</nc>, wherein <nc>information</nc> is passed between <nc>the a key value service and source code artifacts</nc> of <nc>the Design Time Framework</nc> by <nc>way</nc> of <nc>the application program interface</nc> (<nc>API</nc>) of <nc>the Object Service Framework</nc>; and <nc>a Run Time Framework</nc>, running on <nc>a machine</nc> including <nc>a processor</nc>, <nc>that</nc> comprises <nc>pre-built binary code</nc>, wherein in <nc>response</nc> to <nc>an incoming event</nc> for <nc>the web service</nc>, <nc>the Run Time Framework</nc> retrieves from <nc>the second set</nc> of configuration files <nc>a selection</nc> of <nc>the pre-built runtime services</nc> to be placed in <nc>a flow</nc> for <nc>the web service</nc> and <nc>an order</nc> in <nc>which</nc> <nc>the selected pre-built runtime services</nc> will be placed in <nc>the flow</nc>.
| 4
|
4. <nc>The system</nc> of <nc>claim</nc> 1 wherein <nc>the web service</nc> is <nc>a first web service</nc>, <nc>the Design Time Framework</nc> receives <nc>a third set</nc> of <nc>configuration files</nc>, different from <nc>the first set</nc> of <nc>configuration files</nc>, and based on <nc>the third and second sets</nc> of <nc>configuration files</nc>, <nc>the Run Time Framework</nc> generates <nc>a second web service</nc>, different from <nc>the first web service</nc>.
|
8086462
|
10936835
|
1. <nc>A method</nc> of reporting <nc>data</nc> from <nc>an automated spoken dialog service</nc>, <nc>the method</nc> comprising: extracting <nc>data</nc> regarding <nc>user dialogs</nc> with <nc>the automated spoken dialog service</nc>, and not with <nc>a human agent</nc>, using <nc>a dialog logging module</nc> within <nc>the automated spoken dialog service</nc>; analyzing <nc>the data</nc> to identify <nc>trends</nc> for <nc>at least one geographic area</nc> to yield <nc>identified trends</nc>; comparing <nc>the identified trends</nc> for <nc>the at least one geographic area</nc> with <nc>an average</nc> in <nc>a wider geographical area</nc> containing <nc>the at least one geographic area</nc>; modifying <nc>the automated spoken dialog service</nc> on <nc>a geographic basis</nc> according to <nc>the identified trends</nc> and <nc>comparison</nc>; based on <nc>the analysis</nc> of <nc>the data</nc> and <nc>the comparison</nc> of <nc>the identified trends</nc>, generating <nc>top natural language headlines</nc>, <nc>each</nc> of <nc>which</nc> describes <nc>at least one identified trend</nc>; and displaying <nc>the generated top headlines</nc> to <nc>a user</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein analyzing <nc>the data</nc> further comprises analyzing <nc>a time-sequence</nc> associated with <nc>the data</nc>.
|
8244046
|
12301224
|
1. <nc>A non-transitory computer readable medium</nc> having stored thereon <nc>a character string updated degree evaluation program</nc> allowing <nc>a computer</nc> to execute: <nc>an extraction step</nc> of extracting <nc>common part character strings</nc> <nc>each</nc> comprising <nc>successive plural characters</nc> and <nc>each</nc> having <nc>a length</nc> greater than or equal to <nc>an arbitrary threshold value</nc> from <nc>original and updated texts</nc> subjected to <nc>comparison</nc> with each other; a step of arranging <nc>the common part character strings</nc> in <nc>an order</nc> that <nc>the common part character strings</nc> occur in <nc>the respective texts</nc>, to create <nc>divided lines</nc>; and a step of comparing <nc>the divided lines</nc> with each other, calculating <nc>a minimum number</nc> of <nc>substitution</nc> among <nc>the common part character strings</nc> necessary to cause <nc>one divided line</nc> to match <nc>another divided line</nc>, thereby acquiring <nc>a context edit distance</nc>, comprising: <nc>a step</nc> of subtracting 1 from <nc>a number</nc> of <nc>extracted common part character strings</nc> to acquire <nc>an edited point number</nc> as <nc>a number</nc> of <nc>edited points</nc> where <nc>editing</nc> is performed; a step of calculating <nc>a rate</nc> of <nc>a total length</nc> of <nc>remaining character strings</nc> obtained by eliminating <nc>the common part character strings</nc> from <nc>the text</nc> with <nc>respect</nc> to <nc>a total length</nc> of <nc>that text</nc>; <nc>a step</nc> of collecting <nc>remaining character strings</nc> obtained by eliminating <nc>the common part character strings</nc> from <nc>each text</nc> to create <nc>a non-common part character string set</nc>, and <nc>a novelty degree evaluation step</nc> of acquiring <nc>a new creation novelty degree DO</nc> <nc>which</nc> is <nc>a non-matching rate</nc> of <nc>an N-gram</nc> in <nc>a length</nc> less than <nc>the threshold value</nc> among <nc>non-common part character string sets</nc>, wherein <nc>the new creation novelty degree</nc> DO is expressed by <nc>DO=1−|Γ</nc> <nc>1 ∩Γ 2 |/| 1 |</nc> (where <nc>Γ</nc> 1 and <nc>Γ</nc> 2 are <nc>N-gram sets</nc> <nc>each</nc> created from <nc>a non-common part character string set</nc>, <nc>|Γ 1 ∩Γ 2 |</nc> is <nc>a number</nc> of <nc>common elements</nc> commonly occurring in the N-gram sets <nc>Γ</nc> 1 and <nc>Γ</nc> 2 , and <nc>|Γ 1 |</nc> is <nc>a total number</nc> of <nc>N-grams</nc> contained in <nc>the N-gram set</nc> <nc>Γ</nc> 1 ); and a step of calculating <nc>a character string updated degree</nc> from <nc>an evaluation equation</nc>: <nc>a·EP+b·CED+NCP·DO·L</nc> (where <nc>EP</nc> is <nc>a number</nc> of <nc>edited points</nc>, <nc>CED</nc> is <nc>a context edit distance</nc>, <nc>NCP</nc> is <nc>a new creation percentage</nc>, DO is <nc>a new creation novelty degree</nc>, <nc>L</nc> is <nc>a total length</nc> of <nc>an updated text</nc>, and <nc>a</nc> and b are <nc>arbitrary coefficients</nc>), using <nc>the number</nc> of <nc>edited points</nc>, <nc>the context edit distance</nc>, <nc>the new creation percentage</nc>, and <nc>the new creation novelty degree</nc>.
| 6
|
6. <nc>The non-transitory computer readable medium</nc> having stored thereon <nc>the character string updated degree evaluation program</nc> according to <nc>claim</nc> 1 , wherein <nc>the extraction step</nc> comprises <nc>the step</nc> of extracting <nc>a character string</nc> from <nc>the text</nc> as <nc>a common part character string</nc> corresponding to <nc>a portion</nc> where <nc>an offset</nc> of <nc>a dot</nc> plotted on <nc>a dot matrix</nc>, created by comparing <nc>the texts</nc> with each other, from <nc>the center</nc> of <nc>the dot matrix</nc> successively appears as <nc>a constant value</nc> over <nc>predetermined number</nc> of <nc>times</nc> greater than or equal to <nc>the threshold value</nc>.
|
7657832
|
10665156
|
1. <nc>A computer-implemented method</nc> for correcting <nc>an XML electronic document</nc>, <nc>the XML electronic document</nc> having <nc>a structure</nc>, <nc>the method</nc> comprising: identifying <nc>a validation error</nc> in <nc>the XML electronic document structure</nc>, <nc>the validation error</nc> being <nc>an aspect</nc> of <nc>the XML electronic document structure</nc> <nc>that</nc> fails to conform to <nc>rules</nc> of <nc>an XML document type definition</nc> or <nc>an XML schema</nc>, <nc>the rules</nc> being associated with <nc>the XML electronic document</nc>, <nc>the validation error</nc> being of <nc>a particular kind</nc>, wherein identifying <nc>the validation error</nc> includes building <nc>a deterministic finite automation</nc> from <nc>a content model</nc> defined in <nc>a document type definition</nc> of <nc>the XML electronic document</nc> and identifying <nc>the validation error</nc> using <nc>the deterministic finite automaton</nc>; selecting <nc>a suggestion template</nc> from among <nc>multiple suggestion templates</nc> according to <nc>the particular kind</nc> of <nc>the validation error</nc>, and using <nc>the selected suggestion template</nc> to suggest to <nc>a user</nc> suggested <nc>corrections</nc> <nc>that</nc> are predefined in <nc>the template</nc> for <nc>the particular kind</nc> of <nc>validation error</nc>, <nc>the selected suggestion template</nc> including <nc>logic</nc> necessary for modifying <nc>the XML electronic document structure</nc> in <nc>conformance</nc> with <nc>the rules</nc> of <nc>the XML document type definition</nc> or <nc>the XML schema</nc>, wherein modifying <nc>the XML electronic document structure</nc> comprises retagging <nc>an element</nc> in <nc>the XML electronic document structure</nc> and moving <nc>an element</nc> from <nc>a current location</nc> to <nc>a new location</nc> in <nc>the XML electronic document structure</nc>; receiving <nc>an input</nc> selecting one of <nc>the suggested corrections</nc>; and using <nc>the logic</nc> in <nc>the selected suggestion template</nc> to apply <nc>the correction</nc> selected by <nc>the input</nc> to <nc>the XML electronic document</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein suggesting <nc>one or more changes</nc> to <nc>a user</nc> comprises: requesting <nc>information</nc> from <nc>a user</nc> about <nc>the identified structural aspect</nc>; and based on <nc>input</nc> received in <nc>response</nc> to <nc>the request</nc>, suggesting to <nc>the user</nc> <nc>one or more changes</nc> <nc>that</nc> would correct <nc>the identified structural aspect</nc>.
|
10108988
|
14193695
|
1. <nc>A computer-implemented method</nc>, comprising: receiving, at <nc>a primary video content server</nc> including <nc>one or more processors</nc> and <nc>memory</nc>, via <nc>a network</nc>, <nc>a video document</nc>, <nc>the video document</nc> including <nc>video content</nc> to be provided for <nc>presentation</nc> and embedded with <nc>an encoding</nc> specifying <nc>one or more content slot locations</nc>; determining, by <nc>the primary video content server</nc>, based on <nc>the encoding</nc> embedded in <nc>the video content</nc>, the one or more content slot locations at <nc>which</nc> to insert <nc>content slots</nc>, each of <nc>the content slots</nc> configured to receive <nc>content items</nc> for <nc>presentation</nc>, <nc>each</nc> of <nc>the content slot locations</nc> having <nc>a start time</nc> corresponding to <nc>a time</nc> of <nc>the video document</nc> at <nc>which</nc> to insert <nc>the content slot</nc>; inserting, by <nc>the primary video content server</nc>, the content slots at <nc>the determined content slot locations</nc> of <nc>the video document</nc>, <nc>the content slots</nc> configured to receive <nc>content items</nc>; receiving, at <nc>the primary video content server</nc>, from <nc>a client device</nc>, <nc>a request</nc> to provide <nc>the video document</nc> for <nc>presentation</nc>; and delivering, by <nc>the primary video content server</nc>, responsive to receiving <nc>the request</nc> to provide <nc>the video document</nc>, to <nc>the client device</nc>, <nc>the video document</nc> for <nc>presentation</nc> on <nc>the client device</nc>, <nc>receipt</nc> of <nc>the video document</nc> causing <nc>the client device</nc> to: (a) initiate playing <nc>the video document</nc> using <nc>a content rendering application</nc>, (b) execute, subsequent to initiating <nc>playing</nc> of <nc>the video document</nc>, <nc>the encoding</nc> embedded in <nc>the video document</nc> specifying <nc>the one or more content slot locations</nc>, (c) generate, based on executing <nc>the encoding</nc> embedded in <nc>the video document</nc>, <nc>a request</nc> for <nc>supplemental video content</nc> to insert at <nc>the start time</nc> corresponding to <nc>the time</nc> of <nc>the video document</nc> at <nc>which</nc> to insert <nc>the content slot</nc>, and (d) transmit, to <nc>a secondary video content server</nc> including <nc>one or more processors</nc> and <nc>memory</nc>, <nc>the request</nc> for <nc>supplemental video content</nc> to insert in <nc>the content slot</nc>, <nc>receipt</nc> of <nc>the request</nc> for <nc>supplemental video content</nc> causing <nc>the secondary video content server</nc> to<nc>: (i</nc>) select <nc>a content item</nc> from <nc>the plurality</nc> of <nc>content items</nc> to serve in <nc>the content slot</nc> of <nc>the video document</nc> based on <nc>the request</nc> for <nc>supplemental video content</nc> and <nc>(ii</nc>) provide, to <nc>the client device</nc>, responsive to selecting <nc>the video content item</nc>, <nc>the video content item</nc> for <nc>presentation</nc> on <nc>the client device</nc> at <nc>the time</nc> of <nc>the video document</nc> at <nc>which</nc> <nc>the content slot</nc> is inserted; and <nc>(e) insert</nc>, responsive to receiving <nc>the video content item</nc> from <nc>the secondary video content server</nc>, for playing at <nc>the client device</nc>, <nc>the video content item</nc> into <nc>the content slot</nc> specified by <nc>the encoding</nc> embedded in <nc>the video content document</nc>.
| 2
|
2. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , <nc>wherein receipt</nc> of <nc>the video document</nc> causes <nc>the client device</nc> to: analyze <nc>content</nc> of <nc>the video document</nc> to determine <nc>relevancy information</nc> of <nc>the video document</nc>.
|
8219567
|
11080363
|
1. <nc>A method</nc> implemented by <nc>processor-executable instructions</nc> located in <nc>a storage media</nc>, <nc>the method</nc> comprising: determining <nc>a mobile-friendliness indication</nc> <nc>that</nc> is associated with <nc>a uniform resource locator</nc> (<nc>URL</nc>), wherein <nc>the mobile-friendliness indication</nc> indicates <nc>a compatibility</nc> of <nc>a page</nc> associated with <nc>the URL</nc> with <nc>one or more display characteristics</nc> of <nc>a mobile device</nc>; determining <nc>the mobile-friendliness indication</nc> based on <nc>a site comparator</nc>, <nc>a markup language comparator</nc>, <nc>a mobile-specific response examiner</nc>, and <nc>a site content analyzer</nc> to identify <nc>multiple mobile friendliness scores</nc>; identifying <nc>the multiple mobile friendliness scores</nc> based on <nc>a first weight mobile friendliness score</nc> from <nc>the site comparator</nc>, <nc>a second weight mobile friendliness score</nc> from <nc>the markup language comparator</nc>, <nc>a third weight mobile friendliness score</nc> from <nc>the mobile-specific response examiner</nc>, and <nc>a fourth weight mobile friendliness score</nc> from <nc>the site content analyzer</nc>; producing <nc>a weighted average mobile friendliness score</nc> from <nc>the multiple mobile friendliness scores</nc>; in <nc>response</nc> to <nc>the weighted average mobile friendliness score</nc>, determining <nc>the mobile friendliness indication</nc>; and storing <nc>the determined mobile-friendliness indication</nc>.
| 7
|
7. <nc>The method</nc> as recited in <nc>claim</nc> 1 , wherein <nc>the determining</nc> comprises: injecting <nc>a URL-relevant term</nc> <nc>that</nc> is applicable to <nc>a mobile-friendly user agent type</nc> into <nc>the URL</nc> to form <nc>a modified URL</nc>; requesting <nc>a response</nc> from <nc>the URL</nc> using <nc>the modified URL</nc>; examining <nc>the response</nc> received from <nc>the requesting</nc>; and determining <nc>the mobile-friendliness indication</nc> based on <nc>the examining</nc>.
|
9769097
|
14290937
|
1. <nc>A method</nc> for <nc>extensible chat rooms</nc> in <nc>a hosted chat environment</nc>, <nc>the method</nc> comprising: providing <nc>multiple different chat rooms</nc>, each executing in <nc>a separate process address space</nc> of <nc>at least one computing device</nc> in <nc>a hosted chat room environment</nc>, <nc>each</nc> of <nc>the chat rooms</nc> including <nc>a textual transcript</nc> of <nc>chat postings</nc> and <nc>a video feed</nc> of <nc>a subject</nc>; displaying to <nc>different moderators</nc> of <nc>different ones</nc> of <nc>the chat rooms</nc>, <nc>a catalog</nc> of <nc>extensions</nc> to pre<nc>-</nc>process <nc>an event</nc> in <nc>connection</nc> with <nc>the different ones</nc> of <nc>the chat rooms</nc>; responsive to <nc>a selection</nc> of one of <nc>the extensions</nc> by one of <nc>the moderators</nc>, provisioning the selected one of <nc>the extensions</nc> for pre-processing <nc>a particular event</nc> in a corresponding one of <nc>the chat rooms</nc>, <nc>the extension</nc> pre<nc>-</nc>processing <nc>the particular event</nc> as <nc>the particular event</nc> occurs in <nc>the chat room</nc> before permitting <nc>the chat room</nc> to process <nc>the particular event</nc>; and monitoring the selected one of <nc>the extensions</nc> once provisioned in the corresponding one of <nc>the chat room</nc> and terminating <nc>execution</nc> of the selected one of <nc>the extensions</nc> based upon <nc>a monitored behavior</nc> of the selected one of <nc>the extensions</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein the selected one of <nc>the extensions</nc> comprises <nc>an event handler</nc> configured to preprocess <nc>an event</nc> arising from <nc>an entry</nc> of <nc>a participant</nc> to the corresponding one of <nc>the chat rooms</nc>, or <nc>an event</nc> arising from <nc>an egress</nc> of <nc>a participant</nc> from the corresponding one of <nc>the chat rooms</nc>.
|
9129029
|
13111485
|
1. A method comprising: identifying <nc>a region</nc> defining <nc>an area</nc> of <nc>interest</nc> in <nc>response</nc> to receiving <nc>a viewport</nc> from <nc>a client device</nc> corresponding to <nc>the region</nc> including identifying <nc>a plurality</nc> of <nc>content items</nc> <nc>that</nc> are associated with <nc>the region</nc>; evaluating <nc>query logs</nc> associated with <nc>users</nc> <nc>that</nc> submitted <nc>queries</nc> associated with <nc>the region</nc> to determine <nc>a ranking</nc> associated with <nc>the plurality</nc> of <nc>content items</nc>, wherein <nc>each content item</nc> is ranked based on <nc>a number</nc> of <nc>queries</nc> in <nc>the query</nc> logs for <nc>which</nc> <nc>each content item</nc> is responsive and a time of <nc>occurrence</nc> of <nc>the number</nc> of <nc>queries</nc>, wherein <nc>each content item</nc> is ranked higher if <nc>the number</nc> of <nc>queries</nc> occurred more recently on average than <nc>other queries</nc> in <nc>the query</nc> logs and ranked lower if <nc>the number</nc> of <nc>queries</nc> occurred less recently on average than <nc>other queries</nc> in <nc>the query</nc> logs; receiving <nc>a geographic search query</nc> associated with <nc>the region</nc>; and providing one or more of <nc>the content items</nc> in <nc>response</nc> to <nc>the geographic search query</nc> to be displayed on <nc>the viewport</nc> of <nc>the client device</nc>, based at least in <nc>part</nc> on <nc>the ranking</nc> associated with <nc>each</nc> of <nc>the one or more contents items</nc>, wherein <nc>each</nc> of the one or more displayed content items are ranked above <nc>a predetermined threshold ranking</nc>, and wherein <nc>the predetermined threshold ranking</nc> is based on <nc>the viewport</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 where <nc>the content items</nc> include <nc>one or more advertisements</nc>.
|
9092756
|
13330186
|
1. <nc>An information retrieval system</nc>, <nc>the system</nc> comprising: <nc>at least one database</nc> storing <nc>a plurality</nc> of <nc>documents</nc>; <nc>a data structure</nc> caching <nc>a uniform resource locator</nc> for <nc>a document</nc> predefined to be relevant to <nc>at least a portion</nc> of <nc>a taxonomy</nc>; and <nc>a server</nc> coupled to <nc>the at least one database</nc>, <nc>the server</nc> comprising <nc>a processor</nc> and <nc>a memory</nc> storing <nc>instructions</nc> <nc>that</nc>, when executed by <nc>the processor</nc>, cause <nc>the processor</nc> to perform <nc>operations</nc> comprising: receiving <nc>a query</nc> from <nc>an application</nc> presented by <nc>a client access device</nc>; processing <nc>the query</nc> against <nc>the at least one database</nc> to generate <nc>a search result</nc> identifying <nc>at least one document</nc> relevant to <nc>the query</nc> according to <nc>a ranking order</nc>; determining from <nc>the data structure</nc> <nc>the uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant to <nc>the at least a portion</nc> of <nc>a taxonomy</nc> based on <nc>association</nc> of <nc>the query</nc> with <nc>the at least a portion</nc> of <nc>a taxonomy</nc>; processing <nc>the search result</nc> as generated to associate at <nc>a position</nc> of <nc>the ranking order</nc> in <nc>the search</nc> result <nc>the uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant to <nc>the at least a portion</nc> of <nc>a taxonomy</nc> as determined from <nc>the data structure</nc>; and returning <nc>the search result</nc> including <nc>the associated uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant to <nc>the client access device</nc>.
| 7
|
7. <nc>The system</nc> of <nc>claim</nc> 1 , wherein <nc>the operations</nc> further comprise: determining whether <nc>the uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant based on <nc>the at least a portion</nc> of <nc>a taxonomy</nc> cached in <nc>the data structure</nc> exists in <nc>the search result</nc>; and extracting <nc>the uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant from <nc>the search result</nc> if <nc>it</nc> exists before <nc>the uniform resource locator</nc> for <nc>the document</nc> predefined to be relevant is associated with <nc>the search result</nc>.
|
6049339
|
08999381
|
1. <nc>A method</nc> to blend <nc>graphical objects</nc> comprising: obtaining <nc>a page description language representation</nc> of <nc>the graphical objects</nc>, <nc>the graphical objects</nc> having <nc>transparency characteristics</nc> and <nc>color characteristics</nc>; converting <nc>a portion</nc> of <nc>the page description language representation</nc> into <nc>a planar map representation</nc>, <nc>the planar map representation</nc> having <nc>a plurality</nc> of <nc>planar map regions</nc> wherein <nc>each planar map region</nc> is associated with one or more of <nc>the graphical objects</nc>; and assigning <nc>a color</nc> to <nc>at least one planar map region</nc> as <nc>a function</nc> of <nc>the transparency characteristics</nc> and <nc>color characteristics</nc> of <nc>the graphical objects</nc> associated with <nc>the planar map region</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 further comprising sorting <nc>the planar map regions</nc> into <nc>a print order</nc>.
|
7640563
|
10125259
|
1. <nc>A method</nc> for describing <nc>media content</nc> implemented by <nc>a computing device</nc> including <nc>one or more processors</nc>, <nc>the method</nc> comprising: receiving from <nc>a first provider</nc>, <nc>a first description</nc> of <nc>a media content</nc>, wherein <nc>the first description</nc> identifies <nc>a first genre</nc> associated with <nc>the media content</nc>; receiving from <nc>a second provider</nc>, <nc>a second description</nc> of <nc>a media content</nc>, wherein <nc>the second description</nc> identifies <nc>a second genre</nc> associated with <nc>the media content</nc>; determining <nc>a first trust level</nc> associated with <nc>the first provider</nc> and <nc>a second trust level</nc> associated with <nc>the second provider</nc>; utilizing <nc>the one or more processors</nc> for calculating <nc>first and second degree values</nc> based on <nc>the first and second trust levels</nc>, wherein <nc>the first and second degrees</nc> indicate how related <nc>the first and second genres</nc> are to <nc>the media content</nc> relative to one another; and providing for <nc>a user</nc>, <nc>a description</nc> of <nc>the media content</nc>, wherein <nc>the description</nc> of <nc>the media content</nc> comprises <nc>the first and second degrees</nc> and <nc>the corresponding first and second genres</nc>.
| 2
|
2. <nc>The method</nc> as recited in <nc>claim</nc> 1 , wherein <nc>each genre</nc> is selected from <nc>a group</nc> of <nc>genres</nc> consisting of <nc>action</nc>, <nc>adventure</nc>, <nc>horror</nc>, <nc>comedy</nc>, <nc>death</nc>, <nc>mystery</nc>, <nc>police involvement</nc>, <nc>thriller</nc>, <nc>political intrigue</nc>, <nc>romance</nc>, <nc>science fiction</nc>, <nc>period setting</nc>, lives <nc>drama</nc>, <nc>sports interest</nc>, <nc>animal interest</nc>, <nc>medical interest</nc>, <nc>legal interest</nc>, <nc>religious interest</nc>, <nc>historical interest</nc>, <nc>war interest</nc>, <nc>epic production</nc>, <nc>fantasy folklore</nc>, <nc>musical, western, monsters</nc>, <nc>teenage college</nc>, <nc>ethnic interest</nc>, and <nc>soap</nc>.
|
9237211
|
12852481
|
1. <nc>An energy harvesting communication device</nc> configured with <nc>signal booster apparatus</nc>, comprising: <nc>at least a communication apparatus</nc>; <nc>at least an antenna apparatus</nc> communicatively coupled to <nc>the communication apparatus</nc> and in <nc>association</nc> with <nc>at least an input output (IO) device</nc>; <nc>at least a microprocessor</nc> configured with <nc>a software</nc> for controlling <nc>communications</nc> via <nc>the communication apparatus</nc> and for <nc>processing data</nc> associated with <nc>said IO device</nc>; said <nc>at least an antenna apparatus</nc> in <nc>communication</nc> with said <nc>at least a microprocessor</nc>; and at least a sensor apparatus embedded in <nc>silicon substrate</nc> and embedded in <nc>a microfiber material</nc> to provide at least one of <nc>a communication medium</nc>, <nc>communication clarity</nc>, <nc>a detection platform</nc>, <nc>detection selectivity</nc>, and <nc>detection sensitivity</nc>.
| 35
|
35. <nc>The energy harvesting communication device</nc> of <nc>claim</nc> 1 , wherein said <nc>at least a communication apparatus</nc> further configured to analyze <nc>a predetermined threshold value</nc> substantially associated with at least one of: <nc>a software defined radio</nc>; <nc>a structural communication architecture</nc>; <nc>database</nc>; <nc>a compiler</nc>; <nc>a router</nc>; and <nc>an illuminance sensory environment</nc> associated with <nc>voltage</nc> for changing <nc>a screen</nc> into <nc>a mirror environment</nc>.
|
9037467
|
13718241
|
1. <nc>A method</nc> of complementing <nc>a spoken text</nc>, said <nc>method</nc> comprising: receiving <nc>text data representative</nc> of <nc>a natural language text</nc>; receiving <nc>effect control data</nc> comprising <nc>at least one effect control record</nc>, <nc>each effect</nc> control record being associated with <nc>a respective location</nc> in <nc>said natural language text</nc>; receiving <nc>a stream</nc> of <nc>audio data</nc>; analyzing <nc>said stream</nc> of <nc>audio data</nc> as <nc>the stream</nc> of <nc>audio data</nc> is received for <nc>natural language utterances</nc> <nc>that</nc> correlate with <nc>said natural language text</nc> at a respective one of <nc>said locations</nc>; and outputting, in <nc>response</nc> to <nc>a determination</nc> by said analyzing that <nc>a natural language utterance</nc> in <nc>said stream</nc> of <nc>audio data</nc> correlates with a respective one of <nc>said locations</nc><nc>, at least one effect control signal</nc> based on <nc>the effect control record</nc> associated with <nc>the respective location</nc>.
| 6
|
6. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: determining, based on said <nc>effect control data</nc> and in <nc>response</nc> to said <nc>determination</nc> that <nc>a natural language utterance</nc> in <nc>said stream</nc> of <nc>audio data</nc> correlates with a respective one of <nc>said locations</nc>, whether said <nc>respective location</nc> is <nc>an expected next location</nc>, wherein based on said <nc>determining</nc> determines <nc>that</nc> said <nc>respective location</nc> is not said expected next location, said outputting <nc>comprises</nc> outputting, for <nc>at least one running effect</nc>, <nc>a control signal</nc> to stop <nc>the respective running effect</nc>.
|
4099230
|
05601280
|
1. In <nc>a programmed digital computer</nc>, <nc>apparatus</nc> for loading <nc>a program</nc> into <nc>a memory</nc> in <nc>a manner</nc> to be <nc>location</nc> independent so that <nc>each segment</nc> of <nc>the program</nc> may be selectively accessed for <nc>modification</nc> or <nc>debugging</nc> without affecting <nc>the balance</nc> of <nc>the program</nc>, wherein <nc>each given program segment</nc> is identified by <nc>a program segment label</nc> comprised of <nc>a group</nc> of <nc>coded digital signals</nc> <nc>which</nc> precedes <nc>a plurality</nc> of <nc>instructions</nc> <nc>that</nc> follow said <nc>label</nc> and <nc>that</nc> comprise <nc>a program segment</nc>, comprising <nc>a program memory</nc>, <nc>a source</nc> of <nc>a plurality</nc> of <nc>program segments</nc> <nc>each</nc> including <nc>a program segment label</nc> followed by <nc>a plurality</nc> of <nc>instructions</nc>, means for sequentially transferring said plurality of <nc>program segments</nc> into and out of <nc>said program memory</nc>, <nc>a program counter means</nc> for counting <nc>storage addresses</nc> used in providing to said <nc>program memory</nc> the addresses for <nc>storage</nc> and for <nc>read</nc> out of <nc>program labels</nc> and <nc>instructions</nc> from <nc>said program memory</nc>, a label memory means, wherein <nc>each label memory address</nc> is represented by <nc>a program segment label</nc>, <nc>label detecting means</nc> for detecting when <nc>a program segment label</nc> has been transferred by <nc>said means</nc> for transferring said <nc>plurality</nc> of <nc>program segments</nc>, control means, responsive to <nc>a label</nc> being detected by said <nc>label detecting means</nc>, for loading in said <nc>program</nc> counter <nc>the address</nc> at <nc>which</nc> the first instruction after said <nc>detected program label</nc> is stored in <nc>said program memory</nc>, and for addressing said <nc>label memory</nc> means with <nc>the detected program segment label</nc>, and means responsive to said <nc>label detecting means</nc> and said <nc>control</nc> means for storing <nc>said loaded address</nc> in said <nc>label memory</nc> means at <nc>an address</nc> represented by <nc>said detected program segment label</nc>, whereby <nc>a program segment</nc> can be selectively read out for <nc>modification</nc> or <nc>debugging</nc> by addressing said <nc>label memory</nc> means with <nc>a programmed segment label</nc> to obtain <nc>the address</nc> in <nc>said program memory</nc> of <nc>the first instruction</nc> of <nc>the program segment</nc>.
| 8
|
8. <nc>Apparatus</nc> as recited in <nc>claim</nc> 1, wherein said <nc>computer</nc> is in <nc>an execute instruction mode</nc> and there is stored in <nc>said program memory</nc> <nc>an IF instruction</nc> and <nc>its associated label</nc>, and <nc>an associated condition</nc> followed by <nc>a first sequence</nc> of <nc>instructions</nc>, followed by <nc>an ELSEIF instruction</nc> followed by <nc>a second sequence</nc> of <nc>instructions</nc>, followed by <nc>an ELSE instruction</nc>, followed by <nc>a third sequence</nc> of <nc>instructions</nc> followed by <nc>an ENDIF instruction</nc>, wherein said <nc>ELSEIF</nc> has associated <nc>an alternative condition</nc> to <nc>the one</nc> associated with <nc>the IF label</nc> and said <nc>ELSE instruction</nc> immediately procedes <nc>the location</nc> of the first of <nc>the instructions</nc> <nc>which</nc> <nc>the computer</nc> should execute in <nc>the event</nc> that <nc>the conditions</nc> associated with <nc>IF</nc> and <nc>ELSEIF</nc> are not met, said <nc>apparatus</nc> comprising <nc>means</nc> for activating <nc>said program memory</nc> to read <nc>instructions</nc> stored therein from <nc>addresses</nc> provided by <nc>said program counter</nc>, means activated when <nc>an IF instruction</nc> and <nc>its associated label</nc> is read out of <nc>said program memory</nc> for detecting <nc>the condition</nc> associated with said IF <nc>instruction</nc>, <nc>a label stack memory means</nc>, means for activating said <nc>label memory</nc> means to store said IF <nc>label</nc> and <nc>the condition</nc> of <nc>data</nc> in said computer at <nc>the time</nc> said IF <nc>instruction</nc> is read out of <nc>said program memory</nc>, means for comparing <nc>said condition</nc> of <nc>data</nc> stored in said <nc>label stack memory</nc> means with <nc>the condition</nc> associated with said IF <nc>instruction</nc> and producing <nc>a first output</nc> when <nc>they</nc> are identical and <nc>a second output</nc> when <nc>they</nc> are not identical, means responsive to said <nc>first output</nc> to enable said <nc>computer</nc> to execute <nc>the instructions</nc> read out of <nc>said program memory</nc> between said IF <nc>instruction</nc> and said <nc>ELSEIF instruction</nc>, means, responsive to said <nc>ELSEIF</nc> being read out of <nc>said program memory</nc>, when said <nc>computer</nc> has executed said <nc>first sequence</nc> of <nc>instructions</nc> to read said IF <nc>label</nc> out of said <nc>label stack memory means</nc> and to apply <nc>it</nc> as <nc>an address</nc> to said label memory means, means responsive to <nc>the application</nc> of <nc>the IF label</nc> as <nc>an address</nc> to said label memory means, to read into said <nc>program counter</nc> means <nc>the address</nc> stored therein whereby <nc>the next instruction</nc> read from said <nc>program memory</nc> means is the one following <nc>ENDIF</nc>, means responsive to said <nc>second output</nc> to prevent said <nc>computer</nc> from executing <nc>any instructions</nc> until <nc>the ELSEIF Instruction and associated condition</nc> is read from <nc>said program memory</nc>, means activated when <nc>an ELSEIF instruction</nc> and <nc>associated condition</nc> is read from <nc>said program memory</nc> to compare <nc>the condition</nc> of <nc>data</nc> stored in said <nc>label stack memory</nc> means with <nc>the condition</nc> associated with <nc>said ELSEIF</nc> and producing <nc>a third output</nc> when <nc>they</nc> are identical and <nc>a fourth output</nc> when <nc>they</nc> are not identical, means responsive to said <nc>third output</nc> to enable said <nc>computer</nc> to execute <nc>instructions</nc> read out from <nc>said program memory</nc> between <nc>ELSEIF</nc> and <nc>ELSE</nc>, means responsive to <nc>the read</nc> out of <nc>ELSE</nc> from <nc>said program memory</nc> when said <nc>computer</nc> has executed <nc>the instructions</nc> immediately proceeding <nc>it</nc> for reading said IF <nc>label</nc> out of said <nc>label stack memory</nc> means and applying <nc>it</nc> as <nc>an address</nc> to said label memory means, means responsive to said <nc>label memory</nc> means being addressed by said IF <nc>label</nc> to enable said <nc>label memory</nc> means to read into said <nc>program counter</nc> means <nc>the address</nc> stored therein whereby <nc>the next instruction</nc> read out of said <nc>program memory</nc> means is the one following <nc>ENDIF</nc>, means responsive to said <nc>fourth output</nc> to prevent said <nc>computer</nc> from executing <nc>any instructions</nc> read from <nc>said program counter</nc> until <nc>an ELSE instruction</nc> is read from <nc>said program counter</nc>, and means responsive to said <nc>ELSE instruction</nc>, when said <nc>computer</nc> has not executed <nc>the instructions</nc> immediately proceeding <nc>it</nc>, to enable said <nc>computer</nc> to execute <nc>instructions</nc> read from said <nc>program memory</nc> means between said ELSE and said <nc>ENDIF</nc>.
|
10013679
|
15602049
|
1. <nc>A method</nc> comprising: identifying, by <nc>a processor</nc>, <nc>first vehicle service data</nc> represents <nc>terms</nc> of <nc>a natural human language</nc> <nc>that</nc> match <nc>one or more taxonomy terms</nc> within <nc>a defined taxonomy</nc> searchable by <nc>the processor</nc>, wherein <nc>a first identifier</nc> uniquely identifies <nc>the first vehicle service data</nc>; associating, by <nc>the processor</nc>, <nc>a meaning</nc> with <nc>the first vehicle service data</nc> based on <nc>the terms</nc> of <nc>the natural human language</nc> represented by <nc>the first vehicle service data</nc> <nc>that</nc> match <nc>the one or more taxonomy terms</nc>; generating, by <nc>the processor</nc>, first metadata <nc>that</nc> represents <nc>the meaning</nc> associated with <nc>the first vehicle service data</nc>; generating <nc>vehicle service content</nc> based at least in <nc>part</nc> on <nc>the first metadata</nc>; generating, by <nc>the processor</nc>, second metadata <nc>that</nc> represents <nc>a meaning</nc> with <nc>at least a portion</nc> of <nc>second vehicle service data</nc>; aggregating, by <nc>the processor</nc>, at least the first metadata and the second metadata to produce <nc>aggregated metadata</nc>; associating, by <nc>the processor</nc>, the first identifier with <nc>the first metadata</nc>; receiving <nc>a request</nc> for <nc>the vehicle service content</nc>, wherein <nc>the request</nc> includes <nc>the first identifier</nc> and/or <nc>the first metadata</nc>, and in <nc>response</nc> to <nc>the request</nc> for <nc>the vehicle service content</nc>, sending <nc>the vehicle service content</nc> to be displayed by <nc>a service tool</nc>.
| 3
|
3. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the second vehicle service data</nc> comprises <nc>vehicle data</nc> captured from <nc>a vehicle</nc> or <nc>a representation</nc> of <nc>the vehicle data</nc> captured from <nc>the vehicle</nc>, and wherein <nc>the vehicle data</nc> is arranged according to <nc>a communication protocol</nc> used by <nc>a first electronic control unit</nc> in <nc>the vehicle</nc> to communicate with <nc>a second electronic control unit</nc> in <nc>the vehicle</nc> or <nc>a vehicle service tool</nc> connectable to <nc>the vehicle</nc>.
|
9886625
|
15374300
|
1. <nc>A activity recognition robot device</nc> comprising: <nc>a memory</nc> storing <nc>known activity data objects</nc>, wherein <nc>each known activity data object</nc> represents <nc>a known activity</nc> and includes <nc>similarity scoring techniques</nc> and <nc>clustered temporal features</nc>; and <nc>an activity recognition device</nc> coupled with <nc>the memory</nc> having <nc>a processor</nc>, wherein, upon <nc>execution</nc> of <nc>software instructions</nc> stored on <nc>a non-transitory computer readable medium</nc>, <nc>the processor</nc> is configurable to: generate <nc>a plurality</nc> of <nc>temporal features</nc> from <nc>a digital representation</nc> of <nc>an observed action</nc> involving <nc>at least one recognized object</nc> using <nc>at least one feature detection algorithm</nc>; establish <nc>an observed activity data object</nc> comprising <nc>one or more observed temporal feature clusters</nc> generated from <nc>the plurality</nc> of <nc>temporal features</nc>; calculate <nc>a similarity activity score</nc> for <nc>the observed activity data object</nc> relative to at least one of <nc>the known activity data objects</nc> as <nc>a function</nc> of <nc>the similarity scoring techniques</nc> <nc>that</nc> are contextually relevant to <nc>the activity recognition device</nc>, <nc>the clustered temporal features</nc>, and the observed temporal feature clusters; access <nc>an activity recognition results</nc> set as <nc>a function</nc> of <nc>the similarity activity score</nc>; and cause <nc>the robot</nc> to take <nc>action</nc> based on <nc>the activity recognition results</nc> set.
| 19
|
19. <nc>The robot device</nc> of <nc>claim</nc> 1 , wherein the at least one of <nc>the known activity data objects</nc> comprises <nc>domain-specific attributes</nc>.
|
10013504
|
15166832
|
1. <nc>A system</nc> for <nc>search</nc> with <nc>autosuggest</nc>, comprising: <nc>a processor</nc> configured to: determine <nc>a plurality</nc> of <nc>potential query suggestions</nc> for <nc>a partially entered query string</nc>; merge <nc>a plurality</nc> of <nc>categories</nc> associated with <nc>a merchant web site</nc> into <nc>a merged category</nc>, comprising to: determine <nc>a first weight</nc> for <nc>a first category</nc> based on <nc>a first query count</nc> associated with <nc>the first category</nc>; determine <nc>a second weight</nc> for <nc>a second category</nc> based on <nc>a second query count</nc> associated with <nc>the second category</nc>, <nc>the plurality</nc> of <nc>categories</nc> including <nc>the first category</nc> and <nc>the second category</nc>; and merge <nc>the first category</nc> and <nc>the second category</nc> into <nc>a single merged category</nc>, the single merged category being associated with <nc>the merged category</nc>, wherein <nc>the merging</nc> of <nc>the first category</nc> and <nc>the second category</nc> comprises to: aggregate <nc>the first weight</nc> and <nc>the second weight</nc> to obtain <nc>a merged weight</nc>, <nc>the merged weight</nc> being associated with <nc>the merged category</nc>; and automatically suggest <nc>a plurality</nc> of <nc>queries</nc> based on <nc>a query</nc> count for <nc>each</nc> of <nc>the queries</nc>, wherein at least one of <nc>the automatically suggested plurality</nc> of <nc>queries</nc> corresponds to <nc>the merged category</nc>; and <nc>a memory</nc> coupled to <nc>the processor</nc> and configured to provide <nc>the processor</nc> with <nc>instructions</nc>.
| 6
|
6. <nc>The system</nc> recited in <nc>claim</nc> 1 , wherein <nc>the processor</nc> is further configured to: determine <nc>a weight</nc> for <nc>each</nc> of <nc>a plurality</nc> of <nc>potential search query suggestions</nc> based on <nc>the query</nc> count.
|
8965897
|
13407885
|
1. <nc>A method</nc> for improving <nc>the usability</nc> of <nc>product feedback data</nc> comprising: receiving of <nc>a plurality</nc> of <nc>product feedback search parameters</nc> by <nc>an intelligent product feedback analytics tool</nc>, wherein said <nc>plurality</nc> of <nc>product feedback search parameters</nc> pertain to at least one of <nc>a product</nc> and <nc>a group</nc> of <nc>products</nc>; performing <nc>a search</nc> on <nc>plurality</nc> of <nc>product feedback data sources</nc> using <nc>the product feedback search parameters</nc> to gather <nc>product feedback</nc>; obtaining <nc>a plurality</nc> of <nc>product feedback search results</nc> applicable to <nc>the plurality</nc> of <nc>product feedback search parameters</nc>, wherein <nc>each product feedback search result</nc> comprises at least one of <nc>a rating value</nc> upon <nc>a rating scale and feedback content</nc> in <nc>a textual format</nc>; for <nc>each product</nc> represented in <nc>the obtained plurality</nc> of <nc>product feedback search results</nc>, synthesizing <nc>a composite rating value</nc> for <nc>each rating category</nc> of <nc>the rating scale</nc> defined for <nc>the intelligent product feedback analytics tool</nc> from <nc>rating values</nc> contained in <nc>product feedback search results</nc> <nc>that</nc> are applicable to <nc>the product</nc>, wherein <nc>the synthesizing</nc> comprises: converting <nc>the rating value</nc> for <nc>each product feedback search result</nc> to <nc>an equivalent rating value</nc> with <nc>respect</nc> to <nc>the rating scale</nc> defined for <nc>the intelligent product feedback analytics tool</nc>; assigning <nc>each product feedback search result</nc> to <nc>a rating category</nc> of <nc>the rating scale</nc> defined for <nc>the intelligent product feedback analytics tool</nc>, wherein <nc>the converted rating value</nc> of <nc>a product feedback search result</nc> falls within <nc>a rating value range</nc> defined for <nc>the rating category</nc> to <nc>which</nc> <nc>it</nc> is assigned; and expressing <nc>a quantity</nc> of <nc>product feedback search results</nc> assigned to <nc>each rating category</nc> as <nc>a percentage</nc> of <nc>a total quantity</nc> of <nc>product feedback search results</nc> <nc>that</nc> are applicable to <nc>the product</nc>; for <nc>each product</nc> represented in <nc>the obtained plurality</nc> of <nc>product feedback search results</nc>, analyzing <nc>the plurality</nc> of <nc>product feedback search results</nc> for <nc>at least one analytic parameter</nc>, wherein <nc>each analytic parameter</nc> represents <nc>a commonality</nc> among <nc>a subset</nc> of <nc>the product feedback search results</nc> <nc>that</nc> are applicable to <nc>the product</nc>, wherein said <nc>analysis</nc> utilizes <nc>natural language processing techniques</nc>; and presenting <nc>the plurality</nc> of <nc>product feedback search results</nc>, composite rating values, and <nc>the at least one analytic parameter</nc> in <nc>an organized manner</nc> within <nc>a user interface</nc>, wherein <nc>the at least one analytic parameter</nc> presented provides <nc>a context</nc> for <nc>the corresponding composite rating value</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: obtaining <nc>a plurality</nc> of <nc>product feedback search results</nc> applicable to <nc>the plurality</nc> of <nc>product feedback search parameters</nc> in <nc>a non-textual format</nc>, wherein <nc>the product</nc> feedback search results in <nc>a non-textual format</nc> comprises at least one of <nc>audio</nc>, <nc>video</nc> and <nc>image</nc>; and translating <nc>the product feedback search results</nc> into <nc>textual format</nc> to be processed by <nc>the intelligent product feedback analytics tool</nc>.
|
9152672
|
13716560
|
1. <nc>A method</nc> for performing <nc>queries</nc> on <nc>a distributed time series data storage system</nc> having <nc>a time series database</nc> that <nc>stores data blocks</nc> containing <nc>time</nc> stamped <nc>data</nc> across <nc>a plurality</nc> of <nc>computing devices</nc>, and <nc>an index database</nc> that stores <nc>an index</nc> associated with <nc>the time stamped data</nc> in <nc>each data block</nc>, <nc>the method</nc> comprising: sending <nc>a query</nc> to <nc>a query layer</nc> running on <nc>a first computing device</nc>, <nc>the query</nc> specifying <nc>criteria</nc> <nc>that</nc> define <nc>a set</nc> of <nc>data</nc> retrieved from <nc>the time series data storage system</nc> and <nc>an analysis</nc> performed on <nc>the set</nc> of <nc>data</nc>; requesting from <nc>the index database</nc> <nc>the indices</nc> associated with <nc>the data blocks</nc> stored in <nc>the time series database</nc> needed to evaluate <nc>the query</nc>; returning <nc>the indices</nc> back to <nc>the query layer</nc>; preparing <nc>a sub</nc><nc>-</nc><nc>query</nc> <nc>that</nc> produces <nc>appropriate data</nc> matching <nc>the criteria</nc>, <nc>the sub</nc><nc>-</nc><nc>query</nc> including <nc>the criteria</nc> and <nc>a logical operation</nc> performed on <nc>the data</nc> matching <nc>the criteria</nc>; forwarding <nc>the sub</nc><nc>-</nc><nc>query</nc> to <nc>an evaluator</nc> running on <nc>each</nc> of <nc>the plurality</nc> of <nc>computing devices</nc> <nc>that</nc> are identified in <nc>the returned indices</nc> as holding <nc>data</nc> corresponding to <nc>the data blocks</nc> needed to evaluate <nc>the query</nc>; evaluating <nc>the criteria</nc> specified in <nc>the sub</nc><nc>-</nc><nc>query</nc> in <nc>each evaluator</nc> with <nc>respect</nc> to <nc>the data blocks</nc> stored on <nc>the same computing device</nc> on <nc>which</nc> <nc>the evaluator</nc> is running in <nc>order</nc> to select <nc>a subset</nc> of <nc>data</nc>; performing <nc>the logical operation</nc> specified in <nc>the sub</nc><nc>-</nc><nc>query</nc> in <nc>each evaluator</nc> on <nc>the subset</nc> of <nc>data</nc> generated in <nc>that evaluator</nc> in <nc>the evaluating step</nc> above to generate <nc>a sub</nc>-<nc>result</nc>; receiving <nc>each sub</nc><nc>-</nc><nc>result</nc> from <nc>each evaluator</nc> at <nc>an output handler</nc>; and combining <nc>the sub</nc><nc>-</nc><nc>result</nc> from <nc>each evaluator</nc> into <nc>a query result</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 wherein <nc>the time series database</nc> comprises: <nc>a plurality</nc> of <nc>data nodes</nc>, each consisting of one of <nc>the plurality</nc> of <nc>computing devices</nc>, on <nc>which</nc> <nc>specific data blocks</nc> are stored; and <nc>a control node</nc> configured to choose <nc>which data node</nc> will store <nc>each data block</nc> and to record <nc>which</nc> <nc>data blocks</nc> are stored on <nc>each</nc> of <nc>the plurality</nc> of <nc>data nodes</nc>.
|
9767504
|
14070629
|
1. <nc>A computer-implemented method</nc> for determining <nc>alternative product queries</nc>, <nc>the method</nc> comprising: identifying, by <nc>one or more processors</nc>, <nc>a set</nc> of <nc>popular product queries</nc>, including: determining, by <nc>one or more processors</nc>, <nc>a number</nc> of <nc>occurrences</nc> of <nc>a first product model number</nc> of <nc>a first product</nc> in <nc>a plurality</nc> of <nc>prior queries</nc> of <nc>documents</nc> submitted by <nc>various users</nc>; verifying that <nc>the number</nc> of <nc>occurrences</nc> meets <nc>a specified threshold number</nc> of <nc>occurrences</nc>; and storing <nc>the first product model number</nc> in <nc>a database</nc> of <nc>popular product queries</nc> with <nc>model numbers</nc> of <nc>other products</nc>; identifying, by <nc>the one or more processors</nc>, <nc>previous queries</nc> <nc>that</nc> were previously submitted by <nc>other users</nc> and specified <nc>the first product</nc>; identifying, by <nc>the one or more processors</nc> within <nc>the previous queries</nc> <nc>that</nc> were previously submitted by <nc>the other users</nc>, attributes of <nc>the first product</nc> <nc>that</nc> were also specified in <nc>the previous queries</nc>; determining, for <nc>the first product</nc>, a product category <nc>that</nc> includes <nc>products</nc> having <nc>the attributes</nc> of <nc>the first product</nc> <nc>that</nc> were specified in <nc>the previous queries</nc> submitted by <nc>the other users</nc>, including accessing <nc>a database</nc> <nc>that</nc> lists <nc>product categories</nc> and <nc>corresponding attributes</nc> to identify <nc>the product category</nc> using <nc>the attributes</nc> of <nc>the first product</nc> <nc>that</nc> were specified in <nc>the previous queries</nc>; rendering, by <nc>one or more processors</nc> in <nc>a graphical user interface</nc>, <nc>a search field</nc> configured to receive <nc>a search query</nc>; receiving, via <nc>the search field</nc> from <nc>a given user</nc>, <nc>a given query</nc> <nc>that</nc> specifies <nc>a first brand</nc> of <nc>the first product</nc>; selecting, as <nc>a query term</nc> for <nc>the given query</nc>, <nc>the first product model number</nc> from <nc>the set</nc> of <nc>popular product queries</nc> based on <nc>the first brand</nc> being specified by <nc>the given query</nc>; selecting, by <nc>the one or more processors</nc> and in <nc>response</nc> to <nc>receipt</nc> of <nc>the given query</nc>, <nc>a suggested product</nc> <nc>that</nc> has <nc>one or more same attributes</nc> as <nc>products</nc> in <nc>the determined product category</nc> and has <nc>a model number</nc> included in <nc>the set</nc> of <nc>popular product queries</nc>, wherein <nc>the suggested product</nc> is different than <nc>the first product</nc> and has <nc>a model number</nc> <nc>that</nc> differs from <nc>the first product model number</nc> and <nc>a brand</nc> <nc>that</nc> differs from <nc>the first brand</nc>; and in <nc>response</nc> to <nc>the given query</nc> <nc>that</nc> specifies <nc>the first brand</nc> of <nc>the first product</nc>: rendering, by <nc>the one or more processors</nc> within <nc>a first region</nc> of <nc>the graphical user interface</nc>, a search result that visually identifies <nc>the suggested product</nc>; and rendering, by <nc>the one or more processors</nc> within <nc>the first region</nc> of <nc>the graphical user interface</nc>, <nc>one or more search results</nc> <nc>that</nc> visually identify <nc>the first product</nc>, wherein <nc>the one or more search results</nc> <nc>that</nc> visually identify <nc>the first product</nc> include <nc>at least a URL</nc> configured to present <nc>an associated document</nc> corresponding to <nc>the URL</nc> for viewing through <nc>the graphical user interface</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 in <nc>which</nc> selecting <nc>the suggested product</nc> is based on <nc>a number of times documents</nc> containing <nc>the suggested product</nc> are selected by <nc>the various users</nc> from <nc>results</nc> responsive to one or more of <nc>the plurality</nc> of <nc>prior queries</nc>.
|
9753705
|
14666669
|
1. <nc>A method</nc>, comprising: identifying, by <nc>a processor</nc>, <nc>a defined pattern</nc> in <nc>a bytecode</nc> derived from <nc>a high level programming language module</nc>; evaluating, by <nc>the processor</nc>, <nc>a conditional expression</nc> associated with <nc>the defined pattern</nc>, wherein <nc>the conditional expression</nc> comprises <nc>a symbolic name</nc> of <nc>a system property</nc> of <nc>the high level programming language</nc>, <nc>an operator</nc>, and <nc>an identifier</nc> of <nc>an execution platform</nc>; excluding from <nc>a scope</nc> of <nc>compilation</nc>, in <nc>view</nc> of <nc>the evaluating</nc>, <nc>a portion</nc> of <nc>bytecode</nc> associated with <nc>the defined pattern</nc>, wherein <nc>the portion</nc> comprises code exploiting <nc>execution platform-specific capabilities</nc>; and compiling, in <nc>view</nc> of <nc>the scope</nc> of <nc>compilation</nc>, <nc>the bytecode</nc> into <nc>a native code</nc>.
| 4
|
4. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the defined pattern</nc> comprises <nc>an opcode</nc> corresponding to <nc>a high-level programming language construct</nc>.
|
9747274
|
14687155
|
1. <nc>A computer-implemented method</nc> of assessing <nc>similarity</nc> between <nc>character strings</nc> comprising: calculating <nc>an initial similarity score</nc> for <nc>a first character string</nc> and <nc>a second character string</nc> based on <nc>an edit distance algorithm</nc>; identifying <nc>the first character string</nc> and <nc>the second character string</nc> as <nc>candidate similar character strings</nc> from <nc>a data collection</nc> based on <nc>the calculated initial similarity score</nc> being greater than or equal to <nc>a similarity threshold value</nc>; determining, when <nc>the first character string</nc> and <nc>the second character string</nc> are identified as <nc>similar character strings</nc>, <nc>a frequency</nc> of <nc>occurrence</nc> for at least one of <nc>the first character string</nc> and <nc>the second character string</nc> from <nc>a collection</nc> of <nc>character strings</nc> stored in <nc>the data collection</nc>, wherein <nc>the frequency</nc> of <nc>occurrence</nc> comprises <nc>a total number</nc> of <nc>times</nc> that at least one of <nc>the first character string</nc> and <nc>the second character string</nc> is present in <nc>the collection</nc> of <nc>character strings</nc>; and decreasing <nc>an occurrence</nc> of <nc>false designations</nc> of <nc>character strings</nc> as being similar, the decreasing further comprising: adjusting <nc>the initial similarity score</nc> to <nc>a greater value</nc> as <nc>a final similarity score</nc> when <nc>the determined frequency</nc> of <nc>occurrence</nc> is no greater than <nc>a low frequency threshold value</nc>, adjusting <nc>the initial similarity score</nc> to <nc>a lower value</nc> as <nc>the final similarity score</nc> when <nc>the frequency</nc> of <nc>occurrence</nc> is greater than <nc>a high frequency threshold value</nc>, and designating <nc>the first character string</nc> and the second character string as similar based on <nc>the final similarity score</nc> being greater than or equal to <nc>the similarity threshold value</nc>.
| 6
|
6. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , wherein <nc>the data collection stores</nc> <nc>a collection</nc> of <nc>people names</nc> such that <nc>the first character string</nc> and <nc>the second character string</nc> represent <nc>names</nc> of <nc>people</nc>.
|
9330089
|
14934668
|
1. <nc>A method</nc> of automatically managing <nc>a dialogue</nc> with <nc>a user</nc>, <nc>the method</nc> comprising: generating <nc>a generic semantic representation</nc> based upon <nc>received user-input data</nc>, <nc>the generic semantic representation</nc> being independent of <nc>a language</nc> and <nc>an input modality</nc> associated with <nc>the received user-input data</nc>, <nc>the generic semantic representation</nc> comprising at least one of <nc>a list</nc> of <nc>semantic slots</nc> and <nc>a sequence</nc> of <nc>nested semantic slots</nc>; determining <nc>a user-intention</nc> based upon <nc>the received user-input data</nc>, <nc>the generic semantic representation</nc> and at least one of: <nc>a maintained state</nc> of <nc>the dialogue</nc>, <nc>concept data</nc> representing <nc>one or more concepts</nc>, and <nc>history data</nc> representing <nc>history</nc> of <nc>the dialogue</nc>; performing <nc>selection</nc> of <nc>a list</nc> of <nc>data items</nc> based on <nc>any</nc> of <nc>the concept data</nc> representing <nc>the one or more concepts</nc> and attribute <nc>data</nc> representing <nc>one or more attributes</nc> associated with <nc>the generic semantic representation</nc>; and sending <nc>output data</nc> indicative of <nc>one or more actions</nc> for <nc>a dialogue application</nc> to perform, <nc>the one or more actions</nc> being determined based on <nc>a result</nc> of said determining <nc>the user-intention</nc>.
| 8
|
8. <nc>The method</nc> according to <nc>claim</nc> 1 , further comprising updating <nc>the maintained state</nc> of <nc>the dialogue</nc> based on <nc>the data</nc> representing <nc>the one or more attributes</nc>.
|
9691376
|
14962198
|
1. <nc>A method</nc> comprising: synthesizing <nc>speech</nc> from <nc>a text</nc>; identifying <nc>an acoustic unit</nc> <nc>sequential pair</nc> in <nc>the speech</nc>; searching for <nc>a concatenation cost</nc> for <nc>the acoustic unit</nc> <nc>sequential pair</nc> in <nc>a database</nc> using <nc>a hash table</nc> for <nc>the database</nc>; and when <nc>the concatenation cost</nc> is not found in <nc>the database</nc>, assigning <nc>a default value</nc> as <nc>the concatenation cost</nc> for <nc>the acoustic unit</nc> <nc>sequential pair</nc>.
| 7
|
7. <nc>The method</nc> of <nc>claim</nc> 1 , <nc>wherein the database</nc> <nc>stores acoustic units</nc> in <nc>linear predictive coding parameters</nc>.
|
8336021
|
12334838
|
1. At <nc>a computer system</nc> including <nc>one or more processors</nc> and <nc>system memory</nc>, <nc>a method</nc> for formulating <nc>a collection</nc> of <nc>set membership conditions</nc> for <nc>a set</nc>, the method comprising: <nc>an act</nc> of accessing <nc>a set definition</nc> for <nc>a specified set</nc>, <nc>the set definition</nc> defining <nc>what resources</nc> are to be included in <nc>the specified set</nc>; <nc>an act</nc> of accessing <nc>a membership condition grammar</nc>, <nc>the membership condition grammar</nc> generally indicating how to translate <nc>a set definition</nc> for <nc>a set</nc> into <nc>membership conditions</nc> indicative of <nc>membership</nc> in <nc>the set</nc>; <nc>an act</nc> of translating <nc>the accessed set definition</nc> into <nc>one or more corresponding membership conditions</nc> defined in <nc>accordance</nc> with <nc>the membership condition grammar</nc>, <nc>each membership condition</nc> including <nc>one or more membership condition statements</nc>, <nc>each membership condition statement</nc> defining <nc>a condition</nc> about <nc>a resource's attributes</nc> <nc>that</nc> is to be true for <nc>the resource</nc> to be considered for <nc>membership</nc> in <nc>the set</nc>, including for <nc>each membership condition statement</nc>: <nc>an act</nc> of decomposing <nc>a portion</nc> of <nc>the accessed set definition</nc> into <nc>a referent field</nc>, <nc>an attribute field</nc>, <nc>an operator field</nc>, and <nc>a value field</nc> <nc>that</nc> collectively represent <nc>the defined membership condition statement</nc>, <nc>each</nc> of <nc>the referent field</nc>, <nc>attribute field</nc>, <nc>operator field</nc>, and <nc>value field</nc> being defined within <nc>the membership condition grammar</nc> as: <<nc>Referent>::=<Referent</nc> <nc>Reference><Attribute</nc>>, <nc><Attribute>::=a name</nc> of <nc>an attribute</nc> of <nc>a resource</nc> identified by <nc>an expression</nc> to <nc>its left</nc><nc>, <Operator>::=<Relational Operator>|<Inverted Operator</nc>>, and <nc><</nc>Value>::=<Literal <nc>Value>|<Function</nc> <nc>Value>|<De-referenced Value</nc>>, and <nc>further: the attribute field</nc> naming <nc>an attribute</nc>, <nc>the operator field</nc> indicating <nc>a relational operator</nc>, <nc>the value field</nc> representing <nc>a value</nc>, <nc>the referent field</nc> indicating <nc>a referent</nc>, <nc>the referent</nc> referring either directly to <nc>the resource</nc> currently being evaluated for <nc>membership</nc> in <nc>the set</nc> or to <nc>another resource</nc> <nc>that</nc> is related in <nc>some way</nc> to <nc>the resource</nc> <nc>that</nc> is currently being evaluated for <nc>membership</nc> in <nc>the set</nc>; and <nc>an act</nc> of storing <nc>the one or more corresponding membership conditions</nc> for <nc>use</nc> in subsequently determining <nc>resource membership</nc> in <nc>the set</nc>.
| 7
|
7. <nc>The method</nc> as recited in <nc>claim</nc> 1 , wherein <nc>the operator field</nc> indicates <nc>a relational operator</nc> selected from among: <, >, <<nc>=</nc>, =>, <nc>=</nc><nc>=</nc>, and !.
|
8005850
|
11081860
|
1. <nc>A computerized method</nc> for responding to <nc>a user query</nc>, <nc>the method</nc> comprising: receiving <nc>a query</nc> submitted by <nc>a user</nc>; electronically searching, via <nc>a processing device</nc>, <nc>a corpus</nc> comprising <nc>a plurality</nc> of <nc>documents</nc> to identify <nc>one or more hits</nc>, wherein each hit is <nc>a document</nc> from <nc>the corpus</nc> <nc>that</nc> is determined to be relevant to <nc>the query</nc>; generating <nc>a search report</nc> including <nc>a listing</nc> of <nc>the one or more hits</nc>; transmitting <nc>the search report</nc> to <nc>the user</nc>, wherein for <nc>each hit</nc> <nc>that</nc> is not <nc>an annotated hit</nc>, providing <nc>a control element</nc> selectable by <nc>the user</nc> to create <nc>a new annotation</nc> associated with <nc>the hit</nc>; receiving <nc>one or more annotations</nc> related to <nc>the one or more hits</nc> from <nc>the user</nc>, wherein <nc>the one or more annotations</nc> include <nc>one or more user-entered keywords</nc> describing <nc>the one or more hits</nc>; generating <nc>one or more annotation abstracts</nc> for <nc>the one or more hits</nc> based on <nc>the one or more annotations</nc>, <nc>the one or more annotation abstracts</nc> comprising <nc>one or more portions</nc> of <nc>a document</nc> associated with <nc>the one or more hits</nc>, <nc>the portions</nc> of <nc>the document</nc> relevant to <nc>the one or more user-generated keywords</nc> describing <nc>the one or more hits</nc>; adding to <nc>a data store</nc> <nc>the one or more annotations</nc> relating to <nc>the one or more hits</nc> provided by <nc>the user</nc> in <nc>order</nc> to create <nc>a library</nc> of <nc>annotations</nc>; electronically accessing, via <nc>the processing device</nc>, <nc>the library</nc> of <nc>annotations</nc> created by <nc>the user</nc>, <nc>each annotation</nc> being associated with at least one of <nc>the documents</nc> in <nc>the corpus</nc> and including <nc>user-specific metadata</nc> related to <nc>that document</nc>; classifying <nc>the one or more hits</nc> according to <nc>the one or more annotations</nc>; identifying <nc>one or more interests</nc> of <nc>the user</nc> based on <nc>the classification</nc> of <nc>the hits</nc> related to <nc>the one or more annotations</nc>; determining <nc>an intended scope</nc> of <nc>the query</nc> submitted by <nc>the user</nc> based on <nc>the one or more interests</nc>, <nc>the intended scope</nc> comprising <nc>a classification</nc> of <nc>hits</nc> from <nc>a plurality</nc> of <nc>disparate classifications</nc> of <nc>hits</nc>; detecting whether <nc>a search term</nc> associated with <nc>the query</nc> is present for <nc>each document</nc> in <nc>the corpus</nc> associated with <nc>the one or more annotations</nc>, wherein <nc>a given document</nc> is identified as <nc>a hit</nc> in <nc>the event</nc> that <nc>the search term</nc> is present in <nc>the document</nc> and <nc>the document</nc> is associated with <nc>the intended scope</nc> of <nc>the query</nc>, wherein <nc>the user-specific metadata</nc> includes <nc>a plurality</nc> of <nc>fields</nc> and <nc>the query</nc> specifies <nc>which</nc> of <nc>the fields</nc> are to be considered during <nc>an act</nc> of identifying <nc>documents</nc> and <nc>which</nc> of <nc>the fields</nc> are to be excluded during <nc>the act</nc> of identifying <nc>documents</nc>; electronically identifying, via <nc>the processing device</nc>, as <nc>an annotated hit</nc>, <nc>each</nc> of <nc>the hits</nc> <nc>that</nc> is associated with a matching one of <nc>the annotations</nc> in <nc>the fields</nc> being considered and associated with <nc>the intended scope</nc> of <nc>the query</nc>; determining <nc>an inferred rating</nc> for <nc>each hit</nc> <nc>that</nc> is not <nc>an annotated hit</nc> on <nc>the basis</nc> of <nc>similarity</nc> of <nc>a hit</nc> <nc>that</nc> is not <nc>an annotated hit</nc> with <nc>one or more annotated hits</nc> and <nc>a rating</nc> of <nc>the one or more annotated hits</nc>; in <nc>response</nc> to receiving <nc>the one or more annotations</nc>, automatically regenerating <nc>the search report</nc> to update <nc>the search report</nc>, <nc>the search report</nc> comprising: <nc>the listing</nc> of <nc>the hits</nc> ranked accordingly to <nc>the inferred rating</nc> of <nc>the hits</nc> <nc>that</nc> are not <nc>annotated hits</nc> and <nc>the rating</nc> of <nc>the annotated hits</nc>, and <nc>the one or more annotation abstracts</nc>, <nc>the search report</nc> further indicating, for <nc>each hit</nc>, whether <nc>the hit</nc> is <nc>an annotated hit</nc>; and transmitting <nc>the regenerated search report</nc> to <nc>the user</nc>.
| 10
|
10. <nc>The method</nc> of <nc>claim</nc> 1 further comprising: searching <nc>the library</nc> of <nc>annotations</nc> to identify <nc>one or more additional annotated hits</nc>, wherein <nc>each additional annotated hit</nc> corresponds to <nc>a document</nc> from <nc>the corpus</nc> for <nc>which</nc> <nc>the associated annotation</nc> includes <nc>user-specific metadata</nc> <nc>that</nc> is determined to be relevant to <nc>the query</nc>; and incorporating <nc>the additional annotated hits</nc> into <nc>the listing</nc> of <nc>the hits</nc> in <nc>the search results page</nc>.
|
8006268
|
10153346
|
1. <nc>A method</nc> implemented by <nc>a client device</nc> having <nc>a processor</nc> executing <nc>instructions</nc> stored in <nc>computer-readable storage media</nc>, <nc>the method</nc> comprising: receiving <nc>video signals</nc> broadcast on <nc>a multiplexed channel</nc> of <nc>a broadcast network</nc>; extracting from <nc>the received video signals</nc> <nc>a closed captioning stream</nc> of <nc>textual data</nc>; creating <nc>an active list</nc> comprising <nc>a plurality</nc> of <nc>first search terms</nc> by presenting <nc>a plurality</nc> of <nc>questions</nc> at <nc>a user interface</nc> to be answered by <nc>a viewer</nc> to develop <nc>the first search terms</nc> for creating <nc>the active list</nc>; creating <nc>a passive list</nc> comprising <nc>a plurality</nc> of <nc>second search terms</nc> by: monitoring closed captioning <nc>textual data</nc> during <nc>receipt</nc> of one or more previously received <nc>closed captioning streams</nc> of <nc>textual data</nc> of <nc>received video signals</nc> that <nc>the viewer</nc> has viewed or recorded, extracting <nc>words</nc> and <nc>phrases</nc> as <nc>potential search terms</nc> from <nc>the closed captioning textual data</nc>, and automatically selecting <nc>the plurality</nc> of <nc>second search terms</nc> from <nc>the potential search terms</nc> based on <nc>a recentness</nc> and <nc>a frequency</nc> of <nc>occurrence</nc> of <nc>the extracted words</nc> and <nc>phrases</nc>; searching <nc>the stream</nc> of <nc>textual data</nc> for <nc>occurrences</nc> of <nc>textual data</nc> matching one or more of <nc>the first search terms</nc> in <nc>the active list</nc> or one or more of <nc>the second search terms</nc> in <nc>the passive list</nc>, the searching comprising: storing <nc>content programming</nc> corresponding to <nc>the received video signals</nc> in <nc>a buffer</nc>; comparing, by <nc>the processor</nc>, <nc>the closed captioning stream</nc> of <nc>textual data</nc> to <nc>both the active list</nc> and <nc>the passive list</nc>; determining whether <nc>a number</nc> of <nc>matches</nc> of <nc>the first search terms</nc> of <nc>the active list</nc> and <nc>the second search terms</nc> of <nc>the passive list</nc> with <nc>the textual data</nc> exceeds <nc>a threshold number</nc>, wherein <nc>the number</nc> of <nc>matches</nc> is based on <nc>a combination</nc>, over <nc>a period</nc> of <nc>time</nc>, of <nc>a number</nc> of <nc>hits</nc> with <nc>respect</nc> to <nc>the first search terms</nc> in <nc>the active list</nc> and <nc>a number</nc> of <nc>hits</nc> with <nc>respect</nc> to <nc>the second search terms</nc> in <nc>the passive list</nc>; and applying <nc>a greater weight</nc> to <nc>the first search terms</nc> in <nc>the active list</nc> than <nc>a weight</nc> applied to <nc>the second search terms</nc> in <nc>the passive list</nc> when counting <nc>the number</nc> of <nc>matches</nc>; when <nc>the number</nc> of <nc>matches</nc> of <nc>the first search terms</nc> of <nc>the active list</nc> and <nc>the second search terms</nc> of <nc>the passive list</nc> does not exceed <nc>the threshold number</nc> after <nc>a predetermined period</nc> of <nc>time</nc>, ceasing to search <nc>a first closed captioning stream</nc> of <nc>textual data</nc> from <nc>a first channel</nc> before <nc>an end</nc> of <nc>the first closed captioning stream</nc> is reached, deleting <nc>the corresponding content programming</nc> from <nc>the buffer</nc>, and searching instead <nc>a second closed captioning stream</nc> of <nc>textual data</nc> from <nc>a second channel</nc>; and notifying <nc>the viewer</nc> when <nc>the number</nc> of <nc>matches</nc> exceeds <nc>the threshold number</nc> <nc>that</nc> <nc>content programming</nc> determined to be of <nc>interest</nc> to <nc>the viewer</nc> has been located.
| 2
|
2. <nc>The method</nc> according to <nc>claim</nc> 1 , wherein creating <nc>the active list</nc> is further based on <nc>a word omission list</nc> provided by <nc>a third party</nc>.
|
8381119
|
12685623
|
1. <nc>A touchscreen keyboard</nc> for <nc>a pictographic language</nc>, comprising: <nc>a touchscreen display</nc>; and <nc>a display controller</nc> programmed to display <nc>a first arrangement</nc> of <nc>pictographic characters</nc> on <nc>the touchscreen display</nc>, <nc>the first arrangement</nc> comprising <nc>a plurality</nc> of <nc>characters</nc> of <nc>the pictographic language</nc>, wherein <nc>the first arrangement</nc> comprises <nc>a plurality</nc> of <nc>discrete regions</nc> arranged in <nc>a matrix form</nc>, <nc>each</nc> of <nc>the regions</nc> displaying a respective one of <nc>groups</nc> of <nc>characters</nc> selected from <nc>the plurality</nc> of <nc>characters</nc>, <nc>each</nc> of <nc>the regions</nc> including <nc>15 characters</nc> arranged in <nc>3 columns</nc> and <nc>5 rows</nc>, wherein <nc>the display controller</nc> is further programmed to display <nc>characters</nc> in <nc>a matrix form</nc> in a respective one of <nc>the regions</nc>, wherein <nc>characters</nc> in at least one of <nc>the regions</nc> have <nc>the same first phonetic sound</nc> as one <nc>another</nc>, and wherein <nc>the regions</nc> are arranged at <nc>positions</nc> corresponding to <nc>positions</nc> of <nc>keys</nc> of <nc>a QWERTY keyboard</nc>, such that <nc>the first phonetic sounds</nc> of <nc>the characters</nc> of <nc>the regions</nc> substantially correspond to <nc>the phonetic sounds</nc> of <nc>letters</nc> on <nc>the keys</nc> of <nc>the QWERTY keyboard</nc>.
| 14
|
14. <nc>The touchscreen keyboard</nc> of <nc>claim</nc> 1 , further comprising: <nc>a microphone</nc>; and <nc>a speech recognition software program</nc> to recognize <nc>a character</nc> or <nc>word</nc> spoken to <nc>the microphone</nc>, wherein <nc>the display controller</nc> is programmed to highlight <nc>a character</nc> recognized by <nc>the speech recognition software program</nc> on <nc>the touchscreen display</nc>.
|
9773040
|
14703761
|
1. <nc>A program product</nc> comprising <nc>a non-transitory computer readable storage medium</nc> that <nc>stores code</nc> executable by <nc>a processor</nc> to perform: detecting <nc>a search token mnemonic</nc> in <nc>a string</nc>, wherein <nc>the string</nc> comprises <nc>one or more operators</nc> selected from <nc>the group</nc> consisting of <nc>arithmetic operators</nc>, <nc>logical operators</nc>, <nc>text operators</nc>, formatting <nc>operators</nc>, and <nc>date operators</nc> and <nc>the search token mnemonic</nc> is associated to <nc>a search token</nc> comprising <nc>two or more search definitions</nc> and <nc>a balance value</nc>, <nc>each search definition</nc> specifies <nc>a search rule</nc> for <nc>one or more dissimilar search paths</nc> and <nc>that</nc> comprises <nc>a search path</nc>, <nc>a capture label</nc>, <nc>a system output status</nc>, <nc>a search event</nc>, <nc>a history range</nc> <nc>that</nc> specifies <nc>a number</nc> of <nc>historical search results</nc> <nc>that</nc> trail <nc>search results</nc>, <nc>a historical function</nc> <nc>that</nc> is applied to <nc>the search results</nc>, <nc>a result type</nc>, and <nc>a capture tag</nc>; searching <nc>the search paths</nc> using <nc>the search rule</nc>, wherein <nc>the balance value</nc> is in <nc>balance</nc> when <nc>the two or more search definitions</nc> <nc>each</nc> retrieve <nc>search results</nc> in <nc>the search event</nc> and <nc>the search token</nc> is resolved to a Boolean TRUE, <nc>the balance value</nc> is out of <nc>balance</nc> when <nc>the search token</nc> is resolved to <nc>a Boolean FALSE</nc>, and <nc>the search token</nc> is undefined when any one of <nc>the two or more search definitions</nc> does not retrieve <nc>search results</nc>; performing <nc>an operation</nc> specified by <nc>the one or more operators</nc> on <nc>the search results</nc> for <nc>the search token mnemonic</nc>; and replacing <nc>the search token</nc> mnemonic with <nc>a search result</nc> in <nc>the string</nc> in <nc>response</nc> to <nc>the balance value</nc> being in <nc>balance</nc>.
| 11
|
11. <nc>The program product</nc> of <nc>claim</nc> 1 , <nc>the code</nc> further storing <nc>the search result</nc>.
|
7844449
|
11392763
|
1. <nc>A computer-readable medium storing computer-executable instructions</nc> for performing <nc>operations</nc> comprising: clustering <nc>a set</nc> of <nc>data objects</nc> into <nc>a plurality</nc> of <nc>groups</nc>; performing <nc>a first pass</nc> of performing <nc>probabilistic latent semantic analysis</nc> on <nc>the groups</nc>; identifying <nc>a plurality</nc> of <nc>latent classes</nc> of <nc>the set</nc> of <nc>data objects</nc>; calculating <nc>a first conditional probability</nc> of <nc>a data object</nc> of <nc>the set</nc> of <nc>data objects</nc> given <nc>a latent class</nc> of <nc>the plurality</nc> of <nc>latent classes</nc>; estimating <nc>a ranking</nc> of <nc>each latent class</nc>; eliminating <nc>low probability links</nc> between <nc>the set</nc> of <nc>data objects</nc> and <nc>the latent classes</nc> based on <nc>the rankings</nc>, <nc>the low probability links</nc> being determined based on <nc>a predetermined probability threshold</nc>; determining <nc>remaining links</nc> between <nc>the set</nc> of <nc>data objects</nc> and <nc>the latent classes</nc>; performing <nc>a second pass</nc> of <nc>probabilistic latent semantic analysis</nc> on <nc>a result</nc> of <nc>the first pass</nc> based on <nc>the remaining links</nc> between <nc>the set</nc> of <nc>data objects</nc> and <nc>the latent classes</nc>; and calculating <nc>a second conditional probability</nc> of <nc>a data object</nc> of <nc>the set</nc> of <nc>data objects</nc> given <nc>the remaining links</nc> between <nc>the set</nc> of <nc>data objects</nc> and <nc>the latent classes</nc>.
| 5
|
5. <nc>The computer-readable medium</nc> of <nc>claim</nc> 1 , wherein <nc>the estimating</nc> includes determining, for <nc>each</nc> of <nc>the latent classes</nc>, <nc>the ranking</nc> of <nc>that latent class</nc> given a respective one of <nc>the objects</nc>, according to <nc>the following calculation</nc>:
<nc>P</nc> ( <nc>z|d</nc> )= P <nc>( d</nc> ) −1 P <nc>( d|z</nc> ) <nc>P</nc> ( <nc>z</nc> <nc>)∝ P</nc> <nc>( d|z</nc> ) <nc>P</nc> ( <nc>z</nc> ), wherein <nc>z</nc> is <nc>the latent class</nc> and <nc>d</nc> is <nc>the respective given object</nc>.
|
8591560
|
13507853
|
1. <nc>A medical implant assembly comprising</nc>: <nc>a) first and second bone anchors</nc> <nc>each</nc> having <nc>a bone attachment structure</nc> on <nc>one end</nc> and <nc>a channel</nc> at <nc>an opposite end</nc>; b) an elongate core sized and shaped to be received in <nc>the bone anchor channels</nc> and extend <nc>therebetween</nc>; <nc>c) closures</nc> to secure <nc>the core</nc> in <nc>the bone anchor channels</nc>; and wherein d) <nc>the core</nc> is pretensioned between <nc>the bone anchors</nc> so as to elastically elongate while remaining in <nc>tension</nc> between <nc>the bone anchors</nc> during <nc>usage</nc>.
| 5
|
5. <nc>The assembly</nc> according <nc>to claim</nc> 1 including <nc>an elastic sleeve</nc> <nc>that</nc> surrounds <nc>the core</nc> <nc>that</nc> is positioned between <nc>the bone anchors</nc> and compressively engages <nc>both</nc> of <nc>the bone anchors</nc>.
|
8880388
|
13596636
|
1. A system for predicting <nc>a lexical answer types</nc> (<nc>LAT</nc>) in <nc>a question</nc> comprising: <nc>a memory storage device</nc> including <nc>a plurality</nc> of <nc>syntactic frames</nc>; <nc>a processor device</nc> operatively connected to said <nc>memory storage device</nc> and configured to: receive <nc>a question text string</nc>; extract <nc>at least one syntactic frame</nc> from <nc>said question string</nc>, designate, in said <nc>syntactic frame</nc>, <nc>a placeholder</nc> for <nc>an entity</nc> corresponding to <nc>a potential lexical answer type</nc>; and query <nc>a lexical knowledge database</nc> to automatically obtain <nc>at least one replacement term</nc> for <nc>said placeholder</nc> of said <nc>at least one syntactic frame</nc>, wherein said <nc>entity placeholder</nc> is <nc>a part</nc> of <nc>a question focus</nc> indicating <nc>a LAT</nc> of <nc>the question</nc>.
| 6
|
6. <nc>The system</nc> as claimed in <nc>claim</nc> 1 , wherein said <nc>processor device</nc> is further programmed to: <nc>rank</nc> said <nc>one or more replacement terms</nc>; and select <nc>a top-ranked replacement term</nc> as <nc>an inferred lexical answer type</nc> to <nc>said question</nc>.
|
8150883
|
11087918
|
1. <nc>A computer-implemented method</nc> of performing <nc>a predefined operation</nc> on <nc>a data component</nc> for <nc>a specific context</nc>, <nc>the method</nc> comprising: receiving <nc>a user input</nc> specifying <nc>at least one context value</nc>, <nc>the user input</nc> being made for performing <nc>a predefined operation</nc> on <nc>a data component</nc> for <nc>a specific context</nc> characterized by <nc>the at least one context value</nc>, <nc>the data component</nc> being <nc>a context specific representation</nc> of <nc>a context independent data type</nc>; reading <nc>data elements</nc> of <nc>the context independent data type</nc> in <nc>response</nc> to <nc>the user input</nc>, <nc>each</nc> of <nc>the data elements</nc> being (1) configured for defining <nc>semantics</nc> of <nc>data</nc> in <nc>a document</nc> <nc>that</nc> is electronically transmitted between <nc>entities</nc> and (2)_associated with <nc>a harmonization indicator</nc> <nc>that</nc>, when set, causes <nc>the associated data element</nc> to be included in performing <nc>the predefined operation</nc> unless <nc>the context independent data type</nc> explicitly excludes <nc>the associated data element</nc> for <nc>the specific context</nc>, and that, when not set, causes <nc>the associated data element</nc> not to be included in performing <nc>the predefined operation</nc> unless <nc>the context independent data type</nc> explicitly includes <nc>the associated data element</nc> for <nc>the specific context</nc>; and performing <nc>the predefined operation</nc> on <nc>the data component</nc> based at least in <nc>part</nc> on <nc>a setting</nc> of <nc>the harmonization indicator</nc> of <nc>each</nc> of <nc>the data elements</nc>.
| 12
|
12. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 , wherein <nc>the predefined operation</nc> comprises deleting <nc>the data component</nc> for <nc>the specific context</nc>.
|
8965145
|
13729458
|
1. <nc>A dispatcher apparatus</nc> having: <nc>one or more processors</nc>; <nc>a segmenter</nc> stored on <nc>a memory</nc> and executable by <nc>the one or more processors</nc>, <nc>the segmenter</nc> for receiving <nc>an image query</nc> including <nc>an image</nc> and segmenting <nc>the image</nc> into <nc>one or more content-type specific queries</nc>; <nc>a distributor</nc> stored on <nc>the memory</nc> and executable by <nc>the one or more processors</nc>, <nc>the distributor</nc> coupled to <nc>the segmenter</nc> for submitting <nc>the one or more content-type specific queries</nc> to <nc>one or more coupled corresponding content-type index tables</nc> for <nc>recognition</nc>; and <nc>an integrator</nc> stored on <nc>the memory</nc> and executable by <nc>the one or more processors</nc>, <nc>the integrator</nc> for receiving <nc>recognition results</nc> from <nc>the one or more corresponding content-type index tables</nc>, integrating <nc>the recognition results</nc> into <nc>an integrated result</nc> based on <nc>a level</nc> of <nc>agreement</nc> between <nc>the recognition results</nc> and transmitting <nc>the integrated result</nc>.
| 2
|
2. <nc>The dispatcher apparatus</nc> of <nc>claim</nc> 1 , wherein <nc>the one or more content-types</nc> are selected from <nc>a group</nc> of <nc>black text</nc> on <nc>white background</nc>, <nc>black and white natural images</nc>, <nc>color natural images</nc>, <nc>black and white diagrams</nc>, <nc>color diagrams</nc>, <nc>headings</nc>, and <nc>color text</nc>.
|
9411855
|
13162273
|
1. <nc>A method</nc> comprising: receiving <nc>a feed item</nc>, <nc>the feed item</nc> being displayable in <nc>a feed</nc> of <nc>a social networking system</nc> implemented using <nc>a database system</nc>, <nc>the feed</nc> being displayable on <nc>a display device</nc>; processing <nc>textual content</nc> of <nc>the feed item</nc> to detect <nc>a designated keyword</nc> in <nc>the textual content</nc>, <nc>the designated keyword</nc> associated with <nc>a data record creation rule</nc>; responsive to detecting <nc>the designated keyword</nc> in <nc>the textual content</nc> of <nc>the feed item</nc>, automatically: causing <nc>a data record</nc> to be created as <nc>a data object</nc> in <nc>a database</nc> of <nc>the database system</nc>, <nc>the created data record</nc> being accessible via <nc>a cloud-based computing services environment</nc>; identifying <nc>information</nc> of <nc>the feed item</nc> or of <nc>one or more feed items</nc> associated with <nc>the feed item</nc> related to <nc>the created data record</nc>; determining that <nc>the created data record</nc> is related to <nc>a first customer relationship management (CRM) record</nc> of <nc>a CRM system</nc>; and causing <nc>one or more data fields</nc> of <nc>the created data record</nc> to be populated with <nc>the identified information</nc> of <nc>the feed item</nc> and <nc>information</nc> of <nc>the first CRM record</nc>.
| 9
|
9. <nc>The method</nc> recited in <nc>claim</nc> 1 , wherein causing <nc>the created data record</nc> to be created <nc>comprises</nc>: associating <nc>the received feed item</nc> with <nc>the created data record</nc>, <nc>the received feed item</nc> being accessible via <nc>the created data record</nc>.
|
7710590
|
11527934
|
1. <nc>A method</nc> for <nc>generation</nc> of <nc>a print job ticket</nc> comprising: <nc>extracting page</nc> attribute <nc>information</nc> from <nc>a first electronic document</nc>; embedding <nc>the page</nc> attribute <nc>information</nc> as <nc>an object</nc> within <nc>the first electronic document</nc>; extracting <nc>page</nc> attribute <nc>information</nc> from <nc>a second electronic document</nc>; embedding <nc>the page</nc> attribute <nc>information</nc> as <nc>an object</nc> within <nc>the second electronic document</nc>; processing <nc>the first and second electronic documents</nc> to form <nc>a print job</nc>; automatically reading <nc>the objects</nc> embedded in <nc>the processed documents</nc> to retrieve <nc>embedded page</nc> attribute <nc>information</nc>; and generating <nc>a job ticket</nc> for <nc>the print job</nc> based on <nc>the retrieved embedded page</nc> attribute <nc>information</nc> of <nc>the first and second electronic documents</nc>, including automatically assigning a first page attribute to <nc>at least a first page</nc> of <nc>the print job</nc> <nc>which</nc> includes <nc>information</nc> from <nc>the first document</nc> based on <nc>the object</nc> embedded in <nc>the first document</nc> and assigning <nc>a second page</nc> attribute to <nc>at least a second page</nc> of <nc>the print job</nc> <nc>which</nc> includes <nc>information</nc> from <nc>the second document</nc> based on <nc>the object</nc> embedded in <nc>the second document</nc>.
| 2
|
2. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the embedding</nc> includes embedding <nc>the object</nc> in at least one of <nc>the documents</nc> when <nc>the at least one document</nc> is in <nc>an interchange format</nc> <nc>which</nc> is compatible with <nc>an image rendering device</nc> on <nc>which</nc> <nc>the print job</nc> is to be rendered.
|
9483535
|
14229119
|
1. <nc>A computer-implemented method</nc> for expanding <nc>search results</nc>, <nc>at least a portion</nc> of <nc>the method</nc> being performed by <nc>a computing device</nc> comprising <nc>at least one processor</nc>, <nc>the method</nc> comprising: determining that <nc>a user</nc> is attempting to perform <nc>a document search</nc>; presenting <nc>an option</nc> to <nc>the user</nc> to include, within <nc>results</nc> of <nc>the search</nc>, in <nc>addition</nc> to <nc>a found document</nc> <nc>that</nc> matches <nc>search criteria</nc> entered by <nc>the user</nc><nc>, other documents</nc> within <nc>a document family</nc> associated with <nc>the found document</nc>, wherein <nc>the other documents</nc> within <nc>the document family</nc> do not match <nc>the search criteria</nc>; determining that <nc>the user</nc> has selected to include <nc>the other documents</nc> within <nc>the document family</nc> associated with <nc>the found document</nc>; in <nc>response</nc> to determining that <nc>the user</nc> has selected to include <nc>the other documents</nc> within <nc>the document family</nc>, including <nc>the other documents</nc> within <nc>the document family</nc> in <nc>the search results</nc>.
| 8
|
8. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: displaying to <nc>the user</nc> <nc>at least one predetermined category</nc> with <nc>which</nc> <nc>the search results</nc> are associated; after performing <nc>the search</nc>, displaying to <nc>the user</nc> <nc>the number</nc> of <nc>found documents</nc> <nc>that</nc> fall into <nc>the predetermined category</nc>.
|
9621725
|
15135393
|
1. <nc>A computer</nc> implemented <nc>method</nc> for analyzing <nc>chat leakage</nc>, comprising: providing <nc>a processor</nc> configured for obtaining <nc>chat-related information</nc> from <nc>at least one chat session</nc> conducted over <nc>a computer network</nc> via <nc>a chat communications channel</nc> between <nc>a customer</nc> and <nc>an agent</nc>; said <nc>processor</nc> configured for identifying <nc>customer leakage information</nc> from <nc>said chat session</nc> to <nc>another communications channel</nc>; said <nc>processor</nc> configured for building <nc>a model</nc> based on <nc>said chat-related information</nc> and said <nc>leakage information</nc>; said <nc>processor building</nc> said <nc>model</nc> by: using <nc>a chat text</nc> to build <nc>an anchor</nc>; identifying <nc>one or more filters</nc> by extracting <nc>channel names</nc> referred to in <nc>the chat text</nc> <nc>that</nc> is used to build <nc>said anchor</nc> by using <nc>a window</nc> of <nc>words</nc> around <nc>the anchor</nc> to identify <nc>a type</nc> of <nc>channel</nc>; and once <nc>the anchors</nc> and <nc>filters</nc> are identified, using <nc>a priority matrix</nc> to identify <nc>an exact channel</nc>, wherein said <nc>processor</nc> is configured for applying <nc>said model</nc> to identifying <nc>a communications channel</nc> to <nc>which leakage</nc> occurs; and said <nc>processor</nc> configured for summarizing and passing <nc>contextual information</nc> of said <nc>chat session</nc> from said <nc>chat communications channel</nc> to said <nc>another communications channel</nc> to avoid repeating <nc>collection</nc> of <nc>said information</nc> and to allow said <nc>agents</nc> to communicate intuitively with <nc>said customers</nc> to improve <nc>the customer experience</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , further comprising: after said <nc>anchor</nc> is built, using <nc>positive hits</nc> generated during <nc>anchor building</nc> in <nc>connection</nc> with <nc>edit-distance</nc> to obtain <nc>a temporary categorization</nc> of <nc>a team/department</nc>.
|
7734459
|
09899554
|
1. <nc>A computer-implemented method</nc> of associating <nc>dependency structures</nc> from <nc>two different languages</nc> on <nc>a tangible computer</nc> readable medium, wherein <nc>the dependency structures</nc> comprise <nc>nodes</nc> organized in <nc>a hierarchical parent/child structure</nc>, <nc>the computer-implemented method</nc> comprising: associating <nc>nodes</nc> of <nc>the dependency structures</nc> with <nc>a computer</nc> to form <nc>tentative correspondences</nc> on <nc>the tangible medium</nc>, wherein associating includes forming <nc>tentative correspondences</nc> comprising <nc>translations</nc> of <nc>morphological bases</nc> and <nc>derivations</nc>; aligning <nc>nodes</nc> of <nc>the dependency structures</nc> as <nc>a function</nc> of at least one of eliminating at least one of <nc>the tentative correspondences</nc> and <nc>structural considerations</nc> on <nc>the tangible medium</nc>, wherein <nc>aligning</nc> does not require beginning with <nc>either a top or bottom node</nc> of <nc>the hierarchical parent/child structure</nc> of <nc>the dependency structures</nc>; and providing <nc>an output</nc> from <nc>the computer</nc> indicative of <nc>the alignment</nc> of <nc>the dependency structures</nc>.
| 6
|
6. <nc>The computer-implemented method</nc> of <nc>claim</nc> 1 wherein aligning pursuant to <nc>structural considerations</nc> comprises aligning <nc>nodes</nc> as <nc>a function</nc> of <nc>a set</nc> of <nc>rules</nc>.
|
6081814
|
08888464
|
1. <nc>An apparatus</nc> for managing <nc>reference environments</nc>, <nc>the apparatus</nc> comprising: <nc>a computer readable memory device</nc> storing <nc>a plurality</nc> of <nc>reference environments</nc> associated with <nc>a user</nc>, <nc>each reference environment</nc> containing <nc>entries</nc>, <nc>each entry</nc> reflecting <nc>a document</nc>, <nc>the computer readable memory device</nc> storing <nc>data structures</nc> further comprising: <nc>a network directory services system</nc> comprising <nc>directory services objects</nc> operably related in <nc>a hierarchical tree</nc>, <nc>each directory services</nc> object corresponding to <nc>an entity</nc>, and <nc>the hierarchical relationships</nc> corresponding to <nc>the relationships</nc> between <nc>the entities</nc>; <nc>a location object</nc>, being <nc>an instance</nc> of <nc>a directory services object</nc>, containing <nc>reference environment attributes</nc>, and identifying <nc>the reference environments</nc>, for controlling <nc>access</nc> to <nc>the reference environments</nc> by <nc>a user</nc> in <nc>accordance</nc> with <nc>the reference environment</nc> attributes; and <nc>a processor</nc> for executing <nc>a reference environment manager</nc> programmed to create, manage and modify <nc>the plurality</nc> of <nc>reference environments</nc>, <nc>the reference environment manager</nc> being independent from <nc>the directory services system</nc> and effective to create and modify <nc>the entries</nc>, and for executing <nc>a browser</nc> for navigating <nc>the plurality</nc> of <nc>reference environments</nc>.
| 5
|
5. <nc>The apparatus</nc> of <nc>claim</nc> 1, wherein <nc>the reference environment manager</nc> further comprises: <nc>a creator module</nc> for creating <nc>a plurality</nc> of <nc>reference environments</nc>; <nc>a modifier module</nc> for modifying <nc>reference environments</nc>; and <nc>the browser</nc> is further programmed to selectively search, sort, and filter <nc>the reference environments</nc> and <nc>the entries</nc> therein.
|
8127220
|
09734883
|
1. <nc>A computer-implemented method</nc>, comprising: identifying <nc>a document</nc> <nc>that</nc> is stored on <nc>a server</nc> in <nc>a network</nc> and <nc>that</nc> includes <nc>links</nc> to <nc>linked documents</nc>; determining <nc>scores</nc> for <nc>a plurality</nc> of <nc>the links</nc> in <nc>the identified document</nc>; modifying <nc>the identified document</nc> based on <nc>the determined scores</nc>, wherein modifying <nc>the identified document</nc> includes: reordering at least two of <nc>the links</nc> based on <nc>the determined scores</nc>, or sorting at least two of <nc>the links</nc> based on <nc>the determined scores</nc>, comparing <nc>the determined scores</nc> to <nc>a threshold</nc>, and deleting one of <nc>the links</nc> from <nc>the identified document</nc> when <nc>the determined score</nc> for the one of <nc>the links</nc> falls below <nc>the threshold</nc>; and providing <nc>the modified document</nc> to <nc>a user</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein determining <nc>the scores</nc> includes: receiving <nc>input</nc> from <nc>the user</nc>, determining <nc>a score</nc> for <nc>each</nc> of <nc>the linked documents</nc> based on <nc>the received input</nc>, and associating <nc>the determined scores</nc> for <nc>the linked documents</nc> with <nc>the corresponding links</nc> in <nc>the identified document</nc>.
|
9268866
|
14195336
|
1. <nc>A method</nc> of providing <nc>rewards</nc> based on <nc>interactions</nc> with <nc>annotations</nc>, the method being implemented by <nc>a computer system</nc> <nc>that</nc> includes <nc>one or more physical processors</nc> executing <nc>one or more computer program instructions</nc> <nc>which</nc>, when executed, perform <nc>the method</nc>, <nc>the method</nc> comprising: receiving, at <nc>the computer system</nc>, <nc>a first annotation</nc> provided by <nc>a first user</nc> during <nc>a presentation</nc> of <nc>a first content item</nc>, wherein <nc>the first annotation</nc> corresponds to <nc>a reference time</nc> at <nc>which</nc> <nc>a product</nc> or <nc>service</nc> appears in <nc>the first content item</nc>, and wherein <nc>the first annotation</nc> is provided by <nc>the first user</nc> via <nc>a user device</nc> different from <nc>the computer system</nc>; identifying, by <nc>the computer system</nc>, in <nc>the first annotation</nc>, <nc>a reference</nc> referring to <nc>the product</nc> or <nc>service</nc>; modifying, by <nc>the computer system</nc>, <nc>the first annotation</nc> to include <nc>a mechanism</nc> <nc>that</nc> enables <nc>a transaction</nc> related to <nc>the product</nc> or <nc>service</nc>, wherein <nc>the first annotation</nc> is modified to include <nc>the mechanism</nc> based on <nc>the identification</nc> of <nc>the reference</nc>; associating, by <nc>the computer system</nc>, <nc>the modified first annotation</nc> with <nc>a first user account</nc> of <nc>the first user</nc>; providing, by <nc>the computer system</nc>, based on <nc>the corresponding reference time</nc>, <nc>a presentation</nc> of <nc>the modified first annotation</nc> during <nc>one or more presentations</nc> of <nc>the first content item</nc> to <nc>other users</nc> such that <nc>the mechanism</nc> is available for <nc>use</nc> by <nc>the other users</nc> via <nc>the modified first annotation</nc> during <nc>the one or more presentations</nc> of <nc>the first content item</nc>; <nc>monitoring</nc>, by <nc>the computer system</nc>, during <nc>the one or more presentations</nc> of <nc>the first content item</nc>, <nc>use</nc> of <nc>the mechanism</nc> by <nc>the other users</nc> via <nc>the modified first annotation</nc>; and determining, by <nc>the computer system</nc>, <nc>a reward</nc> to be provided to <nc>the first user account</nc> based on <nc>the monitored use</nc> of <nc>the mechanism</nc> by <nc>the other users</nc> via <nc>the modified first annotation</nc>.
| 5
|
5. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the reference</nc> referring to <nc>the product</nc> or <nc>service</nc> comprises <nc>a product or service type identifier</nc> associated with <nc>the product</nc> or <nc>service</nc>, and wherein modifying <nc>the first annotation</nc> comprises supplementing, based on <nc>identification</nc> of <nc>the product</nc> or <nc>service type identifier</nc>, <nc>the product</nc> or <nc>service type identifier</nc> with <nc>the mechanism</nc> in <nc>the first annotation</nc>.
|
9111211
|
13332019
|
1. <nc>A method</nc> for <nc>relevance scoring</nc> of <nc>a digital resource keyword</nc> based on <nc>user actions</nc> associated with <nc>a plurality</nc> of <nc>digital resources</nc>, <nc>the method</nc> comprising: (<nc>a) receiving</nc>, by <nc>a server</nc><nc>, identification</nc> of <nc>a first plurality</nc> of <nc>actions</nc> of <nc>a user</nc>, <nc>each</nc> of <nc>the first plurality</nc> of <nc>actions</nc> comprising <nc>a click</nc> by <nc>the user</nc> on <nc>a link</nc> associated with <nc>a digital resource</nc> of <nc>a plurality</nc> of <nc>digital resources</nc>; <nc>(b) receiving</nc>, by <nc>the server</nc>, <nc>identification</nc> of <nc>a second plurality</nc> of <nc>actions</nc> of <nc>the user</nc> to share <nc>one or more digital resources</nc> of <nc>the plurality</nc> of <nc>digital resources</nc>; (c) identifying, by <nc>the server</nc>, <nc>a plurality</nc> of <nc>keywords</nc> from <nc>content</nc> of <nc>one or more digital resources</nc> of <nc>the plurality</nc> of <nc>digital resources</nc>; (d) classifying, by <nc>the, patterns</nc> from <nc>the first plurality</nc> of <nc>actions</nc>, <nc>the second plurality</nc> of <nc>actions</nc> and <nc>the plurality</nc> of <nc>keywords</nc>; and <nc>(e) generating</nc>, by <nc>the server</nc> based on <nc>the pattern classification</nc>, <nc>a relevance score</nc> responsive to receiving <nc>a user identifier</nc> of <nc>the user</nc> and <nc>a digital resource keyword</nc>, <nc>the relevance score</nc> indicating <nc>a level</nc> of <nc>relevance</nc> between <nc>the digital resource keyword</nc> and <nc>the user</nc>, <nc>the relevance score</nc> based on <nc>the first plurality</nc> of <nc>actions</nc>, <nc>the second plurality</nc> of <nc>actions</nc> and <nc>the plurality</nc> of <nc>digital resources</nc>.
| 11
|
11. <nc>The method</nc> of <nc>claim</nc> 1 , wherein <nc>the digital resource</nc> comprises <nc>any</nc> of the following: <nc>content</nc>, <nc>a web page</nc>, <nc>a uniform resource locator</nc>, <nc>a domain name</nc> and <nc>a phrase</nc>.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.