| Summary: | 碩士 === 國立暨南國際大學 === 資訊管理學系 === 98 === In recent years, rapid development in Internet technology makes the information volume and structural scale of Websites increasing. It becomes more and more important about how to help users find relevant information or services they need more easily. This situation also causes widespread attention about the related technology of Adaptive Web sites. It uses data mining to exploit the websites for analyzing the users' browsing behaviors, based on which the quality of the websites is improved.
Therefore, in this study we propose a Tabu Search algorithm with dynamic tabu tenure to solve the problem about website structure optimization. We will associate different lengths of tabu tenure to various types of moves. In the problem formulation, we add the following features: feasible preceding nodes as an implementation for the compatibility graph constraint, the largest out-degree of pages, the largest in-degree of pages, specified fast-link constraint, page classifications, security considerations and website hierarchy level constraint. These constraints are aiming to make the reorganization of the website structure more conforming to common practice. Finally, we simulate the users’ browsing behavior and perform the experiments with various instances of constraints. The experimental result shows that the proposed algorithm can find a website structure with high usability and logic feasibility.
|
|---|
