Skip to content

ahmaurya/topics_over_time

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 

Repository files navigation

No Maintenance Intended

Topics Over Time

This is an open-source implementation of A Non-Markov Continuous-Time Model of Topical Trends by Xuerui Wang and Andrew McCallum. The paper associated each LDA topic with a beta distribution over timestamps which characterized the evolution of that topic with time.

Instructions

  • Sanitize main_pnas.py and visualize_pnas.py to ensure all input directories, input files, and output directories are present.
  • Run python main_pnas.py to execute Topics over Time algorithm.
  • Run python visualize_pnas.py to visualize the topic-word distributions as well as the beta distributions showing evolution of topics with time.

Dataset

The code is tested on the PNAS titles dataset. The dataset can be found here. The resulting model is pickled and stored in the results folder.

Results

  • Topic Distributions for PNAS Titles Dataset

Topic Distributions

  • Evolution of Topics for PNAS Titles Dataset

Topic Evolution

License

GNU General Public License

Copyright © 2015 Abhinav Maurya

About

Topics over Time implementation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages