How polarizing is a Facebook debate?
McViz is a tool for visualizing the social graph of a Facebook group post. It queries all of the likes and comments from a Facebook thread and constructs a social network. It uses community-detection algorithms to determine polarization (clusters) in arguments and Google's PageRank algorithm to show the most "liked" users.
This project was inspired by a Facebook argument when a friend of ours pointed out that none of the people who liked one side's posts liked the other's and vice versa. McViz attempts to cluster these groups and provide a visual representation.
Below is a sample visualization produced by the program. The vertices correspond to users who participated in a post, the edges correspond to "likes", and the colours represent "communities" detected by the walktrap algorithm.