Zohreh Shokrollahi - Lecturer at Institute of Higher Education ACECR Esfahan
Asghar Karimi - Instructor at Institute of Higher Education ACECR Esfahan

With the rapid development of the World Wide Web and increasing the volume of information, Web research has become an important research area. Web mining research is mainly categorized into two types of web content mining and web usage mining. An important topic in web usage mining is the clustering of users in other words, grouping these users into clusters based on their common features. In this paper, using k-means, Kohonen, and TwoStep methods, we clustered the users into groups with similar characteristics and used the principal component analysis method to enhance clustering quality and use the silhouette criterion to assess clustering quality. Among these three methods, k-means with two clusters had the highest quality and the data set was clustered and analyzed using this method. By analyzing the clusters, can get a better understanding of the users and provide custom and more convenient services for them.

Web mining, Clustering, K-means algorithm, TwoStep algorithm, Kohonen algorithm