Adaptive web site construction using ART

Abstract

The design of a complex web site is a very difficult problem on the World Wide Web. Each user has different goal and sites are usually changing while the site grows and evolves. Web servers record data about user interactions as web log data. This web log has all hits of user access. In order to improve system performance and to reduce visitor's navigation effort, the adaptive web site was suggested. The adaptive web site is a web site that automatically improves its organization and presentation by self-learning. To get information for site modification, the authors use the web log. They cluster the pages in the web site using ART (adaptive resonance theory) and provide grouped pages to improve system performance. The ART model clusters the web log pattern well and the result of clustering information is good to system topology modification.