Chinese Journal of Computers   Full Text
Title	Empirical Study of Region User Behaviors for Web Pages
Authors	MA Wei-Dong1),2) LI You-Ping2) MA Jian-Guo3) ZHOU Ming-Tian1)
Address	1)(School of Computer Science and Engineering, University of Electronics Science and Technology of China, Chengdu 610054) 2)(Institute of Electronic Engineering, China Academy of Engineering Physics, Mianyang, Sichuan 621900) 3)(School of Information Engineering, Southwest University of Science and Technology, Mianyang, Sichuan 621010)
Year	2008
Issue	No.6(960—967)
Abstract &   Background	Abstract The Web-visited bipartite networks, called the user interested networks, display a natural bipartite structure: two kinds of nodes coexist with links only between nodes of different kinds. The Web-visited bipartite networks are constructed dynamically in time series by user requirement behaviors, and the characteristic of user requirement collective behaviors can be analyzed through the bipartite networks. The empirical study of the bipartite networks express that the visiting frequency and in-degree distribution are power-law, which the exponential is between 1.7 and 1.8, and the networks have a clustering characteristics. The bipartite networks can be projected to two kinds of affiliation unipartite networks dividedly from the Web nodes and user nodes, and produce the collective interesting affiliation networks and Web resource affiliation networks. The empirical results express that the edge weight distribution of the unipartite affiliation networks is power-law, and the nodes relations are tightness and clustering. The scale-free and clustering characteristics are important to optimize resource distribution and improve the topology structure and performances of Internet. Keywords complex networks; bipartite networks; scale-free networks; user requirements behaviors Background The work of this paper is supported by the National Natural Science Foundation of China under grant No.60272014 and China Academy of Engineering’ consultation project in 2006. The key problems of the projects are how to improve the information infrastructure of network information environment, construct information share engineering based on Internet and data broadcasting system. This paper presents the statistics characteristics of the Web-visited bipartite networks, and projected to two kinds of unipartite networks: the collective interesting affiliation networks and the Web resource affiliation networks. The networks are dynamically produced in time series that are mainly influence by user requirement collective behaviors. The statistics characteristics of the Web-visited bipartite networks and its affiliation networks in a region network are not reported before the paper. The study results of the scale-free and clustering characteristics for the Web-visited bipartite networks are important to optimize the topology structure of Internet and improve the network performance and resource distribution.