Background
Nowadays Web users are facing the problems of information overload and drowning due to the significant and rapid growth in the amount of information and the large number of users. As a result, how to provide Web users more exactly needed information is becoming a critical issue in Web-based information retrieval and data management. There are a variety of ways that can address the above challenge based on various technical solutions. Among these proposed techniques, Web data mining is devised as an efficient means to discover organizational structure of Web contents, interest-oriented user navigational behaviors and underlying interactions between Web users and Web contents in depth. Web communities are termed as the gathering of Web objects, and could be categorized into Web page community and Web user community dependent on the types of Web objects. The former reflects the functional inherence of Web pages serving, whereas the latter reveals the navigational interest of users, i.e. user community of interest. In this chapter, we mainly aim to address modeling user navigational behaviors (i.e. user communities of interests) and revealing the navigational task space at a semantic level via Web data mining and latent semantic analysis. The proposed chapter will covers a broad range of contents from theoretical aspects such as mathematical models and algorithmic descriptions to experimental investigations including result interpretations, evaluations and comparisons along with an application case study.
Web data mining is a process that discovers the intrinsic relationships among Web data, which are expressed in the forms of textual, linkage or usage information, via analyzing the features of the Web data using data mining techniques. Dependent on the implemented targets, Web data mining consists of Web content mining, Web linkage mining as well as Web usage mining. Particularly, in this chapter we only concentrate our study on discovering Web usage patterns via Web usage mining, and then utilize the discovered usage knowledge for constructing interest-oriented user communities, which could be, in turn, used for presenting Web users more personalized Web contents, i.e. Web recommendation.
Latent Semantic Analysis (LSA) is first proposed in text processing by Deerwester to efficiently tackle the linguistic phenomenon such as synonymy and polysemy. The intuition behind LSA is the concept of “hidden topic”, which is deemed to govern the certainty of co-occurrence of words within the corpus. Recently LSA has been successfully introduced into Web usage mining and has achieved great successes in many related studies e.g. user modeling, collaborative filtering and Web personalization.
In this chapter, we aim to address building user communities of interests via Web usage mining and latent semantic analysis. Meanwhile we also present the application of user communities for Web recommendation.