This repository contains python scripts for detecting community structure in graphs from math publication data. This uses the communities library. The data used is from a previous project as part of the Illinois Geometry Lab.
The folders data/citations and data/papers contain all data.
process_data has scripts which extract similarity matrices from the publication data.
community-detection/community-detect-anim.py creates viz_graph.png and also prints a list of communities detected based on data for common references using the Louvain method.
