Hairball-Buster

Code and data repository for Connections Journal paper on Hairball Buster Graph Analytic Triage method There are three python code files in the attached zip file: • hairball_buster.py, which is the main HB code • hairball_buster_similarity.py, which is the code for producing the y-axis offsets when needed • hairball_buster_first_next_3500.py, which is the code for slicing the data into the temporal graph displays The spreadsheet titled “Pat-export Code DNA sorted” has the data for the CodeDNA dataset with the similarity scores for the Linux coreutils. The first tab is sorted by node, while the second tab is sorted by the similarity score. Two of the other datasets mentioned in the paper are publicly available for download: • Suspended Iranian TwitterTM accounts obtained from https://about.twitter.com/en_us/values/elections-integrity.html#data (TwitterTM, 2018). In this dataset for user-id replies with no retweets between nodes, there were 228,626 nodes and 440,244 edges. • The list of jazz musicians that have performed together with other jazz musicians, thus forming a “link” between them. (Gleiser and Danon, 2003) Community structure in jazz. Advances in Complex Systems, 06(04), 565–573. https://doi.org/10.1142/S0219525903001067