Finding similarity in articles using various clustering techniques

Abstract

Clustering is a vital method within which bunching of articles occurred in the groups in such how that articles of a similar group contain a lot of similarity than the articles into other groups. This paper discussed numerous clustering techniques for finding similarity in articles. These clustering techniques are Hierarchical, K-means, and K-medoids clustering. In this paper, the research focus is to compare several distance measures and find out appropriate distance measure that is used to check the similarity in articles. Distance measure performs a crucial role in the performance of these algorithms. We use different distance measure methods of Hierarchical, K-means, and K-medoids clustering. Here, an experimental examines are performed in Matlab and results show that in Hierarchical clustering Euclidean distance measure, in K-means clustering Correlation distance measure, and in K-medoids clustering City block distance measure provides better results.