期刊名称：International Journal of Mechatronics, Electrical and Computer Technology
印刷版ISSN：2305-0543
出版年度：2015
卷号：5
期号：17
页码：2468-2484
出版社：Austrian E-Journals of Universal Scientific Organization
摘要：Identifying favorite web pages for users has been an important challenge due to the growing web contents. Recently, a number of web personalization systems based on web usage mining (WUM) have been proposed to predict the user-requested web pages. Although these systems provide an automatic tool for analyzing the user's interactions in the Web, the accuracy and coverage of the recommendations should be improved further. In this article, a new hybrid approach based on the web usage mining is proposed for automatic web pages prediction in accordance with the user interests. The proposed system has two phases. In the offline phase, the navigation patterns of users are extracted. To do this, the fuzzy clustering algorithm is used to classify similar user's sessions. This leads to better extraction of the navigation patterns as the user interests have uncertainty. Then, the weighted association rules are extracted for each cluster to reflect the relationships among the pages. In the online phase, the extracted navigation patterns are used to provide users with suitable recommendations. The proposed approach has been evaluated using CTI and NASA datasets. The experimental results show that the proposed approach improves the accuracy and coverage compared to its counterparts
关键词：Personalization; web usage mining; Fuzzy clustering; weighted association rules
