AAAI Publications, The Thirtieth International Flairs Conference

Font Size: 
Fuzzy Spectral Hierarchical Communities in Evolving Political Contribution Networks
Scott Wahl, John Sheppard

Last modified: 2017-05-03


A valuable tool for analyzing social networks is partitioning the complex graphs based on dense sub-networks, usually referred to as communities. This partitioning can discover groups who have similar attributes or behaviors. Using a previously developed method for creating hierarchical fuzzy communities using spectral clustering, these communities are evaluated in a temporal network and tracked through time to evaluate community change. This method is tested on a real world political network based on campaign finance contributions, and it is shown how communities change over time at multiple levels of the hierarchy.


Fuzzy Community; Political Contributions

Full Text: PDF