Extending APAL to Detect Overlapping Communities in Weighted Networks

APAL is an overlapping community detection algorithm that runs on undirected and unweighted graphs of any size. This study extends the algorithm to weighted networks by incorporating the average weight of the community to the evaluation process of the detected communities. Weighted-APAL (WAPAL) is tested with the established LFR benchmark suite against the weighted version of CPM. The resulting normalized mutual information scores show that it can detect weighted communities with a high rate of success.