Semantic Analysis of Web Site Audience by Integrating Web Usage Mining and Web Content Mining

Abstract

With the emergence of the World Wide Web, analyzing and improving Web communication has become essential to adapt the Web content to the visitors’ expectations. Web communication analysis is traditionally performed by Web analytics software, which produce long lists of page-based audience metrics. These results suffer from page synonymy, page polysemy, page temporality, and page volatility. In addition, the metrics contain little semantics and are too detailed to be exploited by organization managers and chief editors, who need summarized and conceptual information to take high-level decisions. To obtain such metrics, we propose a method based on output page mining. Output page mining is a new kind of Web usage mining, between Web usage mining and Web content mining. In our method, we first collect the Web pages output by the Web server. Then, for a given taxonomy covering the Web site knwoledge domain, we aggregate the term weights in the output pages using OLAP tools, in order to obtain topic-based metrics representing the audience of the Web site topics. To demonstrate how our approach solves the cited problems, we compute topic-based metrics with SQL Server OLAP Analysis Service and our prototype WASA for real Web sites. Finally, we compare our results against those obtained with Google Analytics, a popular Web analytics tool.