Web Page Access Prediction Using Fuzzy Clustering by Local Approximation Memberships (flame) Algorithm

Web page prediction is a technique of web usage mining used to predict the next set of web pages that a user may visit based on the knowledge of previously visited web pages. The World Wide Web (WWW) is a popular and interactive medium for publishing the information. While browsing the web, users are visiting many unwanted pages instead of targeted page. The web usage mining techniques are used to solve that problem by analyzing the web usage patterns for a web site. Clustering is a data mining technique used to identify similar access patterns. If mining is done on those patterns, recommendation accuracy will be improved rather than mining dissimilar access patterns. The discovered patterns can be used for better web page access prediction. Here, two different clustering techniques, namely Fuzzy C-Means (FCM) clustering and FLAME clustering algorithms has been investigated to predict the webpage that will be accessed in the future based on the previous action of browsers behavior. The Performance of FLAME clustering algorithm was found to be better than that of fuzzy C-means, fuzzy K-means algorithms and fuzzy self-organizing maps (SOM). It also improves the user browsing time without compromising prediction accuracy.