World Wide Web Usage Mining

World Wide Web data mining includes content mining, hyperlink structure mining, and usage mining. All three approaches attempt to extract knowledge from the Web, produce some useful results from the knowledge extracted, and apply the results to certain real-world problems. The first two apply the data mining techniques to Web page contents and hyperlink structures, respectively. The third approach, Web usage mining (the theme of this article), is the application of data mining techniques to the usage logs of large Web data repositories in order to produce results that can be applied to many practical subjects, such as improving Web sites/pages, making additional topic or product recommendations, user/customer behavior studies, and so forth. This article provides a survey and analysis of current Web usage mining technologies and systems. A Web usage mining system must be able to perform five major functions: (i) data gathering, (ii) data preparation, (iii) navigation pattern discovery, (iv) pattern analysis and visualization, and (v) pattern applications. Many Web usage mining technologies have been proposed, and each technology employs a different approach. This article first describes a generalized Web usage mining system, which includes five individual functions. Each system function is then explained and analyzed in detail. Related surveys of Web usage mining techniques also can be found in Hu, et al. (2003) and Kosala and Blockeel (2000).