Towards Causal Explanations of Community Detection in Networks

摘要

Community detection is a significant research problem in Network Science since it identifies groups of nodes that may have certain functional importance - termed communities. Our goal is to further study this problem from a different perspective related to the questions of the cause of belongingness to a community. To this end, we apply the framework of causality and responsibility developed by Halpern and Pearl [11]. We provide an algorithm-semi-agnostic framework for computing causes and responsibility of belongingness to a community. To the best of the authors' knowledge, this is the first work that examines causality in community detection. Furthermore, the proposed framework is easily adaptable to be also used in other network processing operations apart from community detection.