Topic model

The input for this page is the summary data ("abstracts") of each item in the study carrel. Use this page to first enumerate sets of latent themes ("topics"). Then use the other functions of the page to visualize the results, refine the themes, and download files for offline analysis. This page's functionality is wholly derived from David Mimno's Javascript library called "jsLDA".

Iterations: 0 Train with 20 topics
  • Topic Documents
  • Topic Correlations
  • Time Series
  • Downloads
  • Vocabulary
Documents are sorted by their proportion of the currently selected topic, biased to prefer longer documents.
Words occurring in only one topic have specificity 1.0, words evenly distributed among all topics have specificity 0.0.
WordFrequencyTopic SpecificityStoplist
Documents are grouped by their "date" field (the second column in the input file). These plots show the average document proportion of each topic at each date value. Date values are not parsed, but simply sorted in the order they appear in the input file.
Topics that occur together more than expected are blue, topics that occur together less than expected are red.
Each file is in comma-separated format.