Fuzzy Document Clustering based on Frequent Features and Feature Length

Abstract

Document Clustering is a method of grouping similar documents into one cluster. Fuzzy document clustering is a soft computing technique used for clustering the similar documents. It permits overlapping i.e. it permits single document to belong to multiple clusters. Feature selection and feature extraction is the most important phase during clustering process. In the Literature different feature reduction methods are proposed. In this research paper we have proposed a feature reduction method based on feature frequency and feature length. In this method, we have chosen the features based on no. of occurrence in a set of N documents. We have also taken into account feature length. Finally we have applied fuzzy C-Means clustering algorithm for clustering the N documents into K-Clusters.