Adaptive Site Design Based on Web Mining and Topology

Considering the deficiencies of current Web site in adaptive ability, especially the lack of analysis to its own site topology, this paper, from the user´s view, combines Web log mining with site topology analysis to present the concept of landmark coefficient to measure the significance of Web site nodes. Subsequently, the process for designing adaptive Web site is proposed based on landmark coefficient. Then the adaptive strategies for Website are presented including intelligent recommendation (such as highlight and dynamic map) and buffering for prefetching. Finally, this paper sets up a prototype system for experimental analysis after adopting adaptive strategies. The results show that the implementation of adaptive strategies can improve the users visiting efficiency with shorter length of affair path and less visiting times to meddle nodes.