Structural Analysis of Web User Communities(Web Data Mining)(Joint Workshop of Vietnamese Society of AI, SIGKBS-JSAI, ICS-IPSJ, and IEICE-SIGAI on Active Mining)

概要

論文の詳細を見る
There are two kinds of communities in the Web; communities of related Web pages (Web communities) and communities of users who watch such related pages (user communities). Discovery of the former communities has been attempted by many researchers such as Ku-mar's trawling and Flake's method. Discovery of the latter communities is also important for clarifying the behaviors of Web users. Moreover, it is expected that the characteristics of user communities in the Web correspond to those in real human communities. The author pro-posed a method for discovering user communities based on client-level log data. Web audience measurement data are used as the description of users' Web watching behaviors. Maximal complete bipartite graphs are searched from the graph obtained from the log data without analyzing the contents of Web pages. Since there are many user communities discovered in the above method, choosing a small number of "interesting" ones is required. As the criteria for judging interestingness of user communities, discrepancies of distance among community members are proposed in this paper.
一般社団法人電子情報通信学会の論文
2004-11-28