Update README.md

README.md

@@ -88,7 +88,7 @@ Our TOMT-KIS dataset is available in JSONL format. For each question, we include
   "author": "alany611",
   "url": "http://www.reddit.com/r/tipofmytongue/comments/2gbnla/tomt_1990s_educational_cartoon_for_kids_to_learn/",
   "permalink": "/r/tipofmytongue/comments/2gbnla/tomt_1990s_educational_cartoon_for_kids_to_learn/",
-  "title": "[TOMT] 1990's Educational Cartoon for kids to learn French",
+  "title": "1990's Educational Cartoon for kids to learn French",
   "content": "Hi all,\n\nWhen I was really young, 3-5, I remember watching a cartoon that I think was supposed to teach kids French. I would guess it was made from 1990-1995, but possibly earlier.\n\nIt was in color and the episodes I remember featured a guy with a long, narrow, and crooked nose and greenish skin teaching kids how to count? There was also a scene that had some character running up a clock tower to change the time.\n\nOverall, it was a pretty gloomy feel, iirc, and I'd love to see it again if possible.",
   "created_utc": "1410647042",
   "link_flair_text": "Solved",
@@ -119,7 +119,8 @@ Our TOMT-KIS dataset is available in JSONL format. For each question, we include
   "chosen_answer": "Muzzy?\n\nhttps://www.youtube.com/watch?v=mD9i39GENWU",
   "links_on_answer_path": [
     "https://www.youtube.com/watch?v=mD9i39GENWU"
-  ]
+  ],
+  "categories": [ "[TOMT]" ]
 }
 
 ```
@@ -127,18 +128,19 @@ Our TOMT-KIS dataset is available in JSONL format. For each question, we include
 
 ## Data Fields
 
-Overall, TOMT-KIS includes 128 attributes for each question, such as
+Overall, TOMT-KIS includes 129 attributes for each question, such as
 
 - `id` (string): unique Reddit-identifier of a question
-- `title` (string): the title of a question
+- `title` (string): the title of a question without category tags
 - `content` (string): the main text content of a question
 - `created_utc` (string): the posted question's timestamp
 - `link_flair_text` (string): indicates whether the question is solved; set by moderators
 - `comments` (string, json): the complete tree of the discussion on each question
+- `categories` (list of strings): list of all category tags that were part of the original title
 
 We run our precision-oriented answer identification heuristic on questions tagged as solved by a moderator and add four "new" attributes when the heuristic could identify an answer:
 
-- `answer_detected` (string): indicates whether our heuristic could extract an answer
+- `answer_detected` (bool): indicates whether our heuristic could extract an answer
 - `solved_utc`(string): the timestamp when the identified answer was posted
 - `chosen_answer`(string): the extracted answer
-- `links_on_answer_path`(string in list format): contains all Reddit-external pages that were found in posts between the question and the post with the answer.
+- `links_on_answer_path`(list of strings): contains all Reddit-external pages that were found in posts between the question and the post with the answer.