Optimal K-Means Clustering Algorithm for Weblog Mining

Abstract

World wide web (WWW) generates a huge number of unstructured data and information. The information is stored in weblog file. Weblogs information can be analyzed and visualized by various clustering algorithm. In this work, the k-means clustering algorithm is applied for grouping of the users with similar interest based on accessing of similer information. The optimal value of k is also determined by Elbow method to obtained optimal numbers of clusters. Clustering results are analyzed by various values of k. Comparative analysis ofvarious methods used for selecting the optimal number of clusters are also analyzed.