Akademik Veri Yönetim Sistemi
9th Workshop on Intelligent Techniques for Web Personalization and Recommender Systems, ITWP 2011 - In Conjunction with the 22nd International Joint Conference on Artificial Intelligence, IJCAI 2011, Barcelona, İspanya, 16 Temmuz 2011, cilt.756, ss.57-64
The World Wide Web is an endless source of information, which is mainly represented in the form of Web pages. The way that the users browse the Web depends on both user-oriented factors like the information the users are seeking for and site-oriented factors like the attractiveness of the Web sites, the structure or more specifically, the navigational organization. However, the site's design changes through time, due to factors like including more information about items or introducing new items. In addition, the user's needs and preferences also change, which together bring difficulties for building Web sites that best suit users' needs. Therefore, it is a very important and challenging task to adapt the Web sites automatically in order to facilitate users' navigation in the Web site. This paper proposes a solution to adapt Web sites structural organization according to users' navigation patterns. In the proposed solution, the adaptive web problem is formulated as a classical AI search problem, and a novel Hill Climbing based solution is devised. The proposed solution is realized in a framework, namely IntWEB. The technique is applied to a real-life case study and results are discussed in order to evaluate IntWEB's performance.