This outlines the steps we have taken to create bipartite projections using COP tweets
- Create a subset of each COP dataset based on the normative topics extracted from BERTopic
- Create bipartite network files containing the twitter auther username (source) and topic they interacted with (target) using the
Bipartite.ipynb
script - Create user projections of the bipartite network using the
projs.py
script. Due to the sizes of the files, we used the ARC HPC withjob_script1.sh
.