Title: SVD-initialised K-means clustering for collaborative filtering recommender systems

Authors: Murchhana Tripathy; Santilata Champati; Srikanta Patnaik

Addresses: Information Systems and Technology, T A Pai Management Institute, Manipal, Karnataka, India ' Department of Mathematics, Institute of Technical Education and Research, Siksha 'O' Anusandhan Deemed to be University, Bhubaneswar, Odisha, India ' Department of Computer Science and Engineering, Institute of Technical Education and Research, Siksha 'O' Anusandhan Deemed to be University, Bhubaneswar, Odisha, India

Abstract: K-means is a popular partitional clustering algorithm used by collaborative filtering recommender systems. However, the clustering quality depends on the value of K and the initial centroid points and consequently research efforts have instituted many new methods and algorithms to address this problem. Singular value decomposition (SVD) is a popular matrix factorisation technique that can discover natural clusters in a data matrix. We use this potential of SVD to solve the K-means initialisation problem. After finding the clusters, they are further refined by using the rank of the matrix and the within-cluster distance. The use of SVD based initialisation for K-means helps to retain the cluster quality and the cluster initialisation process gets automated.

Keywords: recommender systems; collaborative filtering; singular value decomposition; SVD; K-means initialisation; within-cluster distance; rank of the matrix.

DOI: 10.1504/IJMDM.2022.119580

International Journal of Management and Decision Making, 2022 Vol.21 No.1, pp.71 - 91

Received: 05 Jan 2021
Accepted: 23 Feb 2021
Published online: 09 Dec 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article