Overlapping community detection model in collaborative networks

Community detection is an important task in complex network analysis. A community (or cluster) is a set of nodes that have more connections inside the set than outside. In the most real social networks, nodes are shared among different clusters and form overlapping communities. In this paper, we introduce a fuzzy graph clustering model to address the overlapping community detection problem. The new model is developed based on the cluster center uncertainty and detects duocentered communities. The interval type-2 fuzzy numbers are used to describe the nodes´ degree of belonging to the communities. We apply this approach on co-authorships network of NAFIPS conferences to determine collaborations and evolving communities.