異なる視点からのノード機能に基づくコミュニティ抽出法

卓恭 伏見, 和巳 斉藤, 哲夫 池田 et al. · Institutional Repositories DataBase (IRDB) · 2013

In this paper, we attempt to extract functional community based on personalized view of each node.A conventional method of functional community extraction can divide all nodes into some node groups with similar global functions based on objective viewpoint for whole of a network.On the other hand, we could assume that functions and roles of each node can differ according to viewpoint shifts.Thus we propose a concept of personalized functional community and a method of extracting these communities based on viewpoint of each node(ego node).Concretely, we adjust an initialized vector of PageRank and introduce an ego-jump mechanism to the proposed method.From experimental results using an artificial and two real networks, we show that our proposed method can divide neighbor nodes from the ego node into fine clusters and distant nodes into coarse clusters, and also divide functionally similar nodes based on objective viewpoint into several clusters according to relation with the ego node.

Cited by 0 works. Open-access PDF

Authors