An intelligent similarity measure for effective text document clustering

Abstract

In many applications, continuous use of multiple queries has become the focal point of many researches. Various evaluations on document clustering methods reveal the need for effective methodology to capture the actual requirement of the user from the web browsers. An intelligent similarity measure with the concept of Neural Network algorithm has been proposed. Experiments show that application of Echo State Neural Network and Radial Basis Function to the training data set gives better clustering of text documents based on the stored weights in order to avoiding retrieval of irrelevant documents.