An Approach to Find Related Communities Based on Bipartite Graphs

In this paper we investigate the problem of extracting related community information from a large collection of Web-pages by performing hyperlink analysis. We consider a group of communities (say related community set) related if they have common interests on some topic. In the proposed approach, we employ dense bipartite graph (DBG) abstraction for two purposes. From the given page collection, we first extract all the communities by mathematically abstracting the community as a DBG over a set of pages. Next, our approach extracts related communities among the communities by abstracting related community set as a DBG over set of communities. We report experimental results on 10 GB TREC (Text REtrieval Conference) data collection that contains 1.7 million pages and 21.5 million links. The results demonstrate that the proposed approach extracts related community structures. Index terms Web mining; Communities; trawling; Link analysis