Topic Modeling GUI


Stage: Active

Campaign: Tool Feedback

a web application for exploring and visualizing topics discovered within text content using the statistical method of topic modeling (possibly a wrapper for Mallet - - a common topic modeling tool). Aimed at technical novices unable to use Mallet’s command-line tools.


Submitted by

Feedback Score

19 votes
Voting Disabled

Idea Details

Vote Activity (latest 20 votes)

  1. Agreed
  2. Agreed
  3. Disagreed
  4. Agreed
  5. Agreed
  6. Agreed
  7. Agreed
  8. Agreed
  9. Agreed
  10. Disagreed
  11. Disagreed
  12. Agreed
  13. Agreed
  14. Agreed
  15. Agreed
  16. Agreed
  17. Agreed
  18. Agreed
  19. Agreed
  20. Disagreed
(latest 20 votes)

Similar Ideas [ 4 ]


  1. The idea was posted


  1. Comment
    Shawn Graham

    There's this already: which works quite well (wrapper for Mallet) though it could use improvement with more parameters, diagnostics, maybe also output where composition is in a matrix of docs by topics. Anyway, the tool does nice HTML output of the results, but maybe some more visualizations would be nice...?

  2. Comment
    Tim McCormick

    A possible application of or related project to this is automatic hashtagging of tweets based on topic analysis of tweet and/or target text. One use of that is to allow users much better, fine-grained tuning of trackers and filters based on hashtags. I suggested this concept of "signal-honing" by hashtag in a Twitter discussion at

  3. Comment

    Good visualization could really increase the usefulness of topic modeling. Great Idea.

  4. Comment
    Natalia Cecire

    Is the obstacle to topic modeling that it's not user-friendly enough, or is it that we still to this day do not have a persuasive theorization of what a topic actually is? I think it's the second one. If y'all run a "one week one theoretical problem," I'd vote for "what in FSM's tasty marinara-covered name is a topic."

Add your comment