中文摘要：

社区探测是图和网络领域非常关键的技术之一,其中聚类方法扮演了重要的角色。针对层次聚类算法较高的时间复杂度,在信息理论框架下提出一种改进的社区探测方法 pIBD。pIBD把单部网络变换成二部图网络,预测k值,并基于信息瓶颈理论进行划分式聚类。实验结果表明,p IBD方法可以获得较已有层次聚类方法更高的准确率。

英文摘要：

Community detection is one of crucial techniques in graph and network research, where clustering plays an important role. Taking into account high time complexity of hierarchical clustering, an improved com- munity detection method, called pIBD, is proposed under information-theoretic framework. The pIBD trans- forms a unipartite network into a bipartite network, predicts the value of k, and implements partitional cluste- ring based on information bottleneck theory. Experimental results show that pIBD could achieve higher accura- cy than existing hierarchical clustering methods.