This is a tool for exploring topic models built on top of streamlit.io. It requires gensim, graphviz, pyvis, networkx, and nltk.

There are two versions of the tool with different objectives:

Supports the workflow to create topic co-occurrence networks and keyword co-occurrence networks. To use this tool, run:

streamlit run tme.py

To learn more about the approach behind the tool you can watch the video of my talk on Exploratory literature reviews using topic modeling at Networks 2021.

Creates a synthesis of multiple runs of a topic model. This is important when using short documents, as there can be quite some variation between individual runs.

streamlit run tme-s.py