The Study Of Web User Fuzzy Clustering Based On Path Similarity

Web usage mining is the process of finding usage patterns from web log data byemploying data mining technology.This paper proposes a web user fuzzy clustering algorithm based on pathsimilarity. Two calculation methods of the degree of user interest are given in thealgorithm, the most interesting pages of user access paths are extracted by the degreeof user interest, then users are clustered according to the page access path similarity.Experimental results show that this method can reduce computation, improve theefficiencyof clustering, and get a satisfactorycollection of user clustering.