/Topic-Modeling-for-Citation-Statements

An overview of topic modeling for citation statements to identify groups of citations and improve topic representations for those

Primary LanguageJupyter Notebook

Topic Modeling for Citation statements

This repo goes over several methods for topic modeling for citation statements. Including:

  • LDA (LDA.ipynb)
  • Contextual Topic Models (CTM.ipynb)
  • Zeroshot Topic Models (ZTM.ipynb)
  • Multi-document and Query Gen for Better Topic Representations (MDS.ipynb)

This repo uses a set of more than 700 citations to (10.1007/s11192-015-1765-5) to identify groups of citations and provide human readable representations. Other than LDA, a GPU should be used for the other notebooks.