Clustering Fuzzy Web Transactions with Rough k-Means

Time duration and presence of a Web page are two factors disclosing Web users' interest. The time duration on a web page is characterized as a fuzzy linguistic variable because it is easily understandable for people and the subtle difference between two durations is disregarded. Thus a Web access pattern is transformed as a fuzzy Web access pattern, which is a fuzzy vector that are composed of n fuzzy linguistic variable or 0. Furthermore, the clusters in Web access patterns do not necessarily have crisp boundaries. This paper proposes a modified k-means clustering algorithm based on properties of rough set to group the gained fuzzy Web access patterns. Finally, an example is provided for clustering the given Web access patterns. The results are proved to be effective.