基于引力度扩展的重叠社区发现算法

Overlapping community detection algorithm based on expansion of gravitational degree

网络中存在社区,而社区之间又出现了重叠节点,即重叠社区。为了发现复杂网络中的重叠社区,在种子扩展与牛顿万有引力定律思想的基础上,提出了一种基于引力度的种子选取策略的重叠社区发现算法。该算法主要以引力度最大的节点作为种子来找到初始社区,根据社区隶属度值选择性添加节点来进行社区扩展。将算法在真实网络上进行测试,实验结果表明,该算法能够有效的挖掘出存在于网络中的重叠社区。

The overlapping community that there are a lot of overlapping nodes during communities in networks. To find overlap- ping communities in complex networks, an algorithm of overlapping community detection that is based on seed expansion and the Newtonian gravitation is proposed. The algorithm whose main strategies are to find an initial partial community from a node with maximal gravitational degree and add nodes to expand the partial community according to the value of community belonging de- gree. The applications to some real-world complex networks to evaluate the algorithm, and the experimental results demonstrate that the algorithm is efficient for detecting overlapping communities in complex networks.