Personal Recommendation via Modified Collaborative Filtering

TL;DR

This paper introduces a modified collaborative filtering method that improves recommendation accuracy by incorporating node degree influence into similarity calculations in bipartite networks.

Contribution

It presents a novel similarity measure for bipartite networks and demonstrates its effectiveness in enhancing collaborative filtering accuracy.

Findings

Significant accuracy improvement over standard methods

Effective for both user-based and object-based filtering

Validated on a benchmark database

Abstract

In this paper, we propose a novel method to compute the similarity between congeneric nodes in bipartite networks. Different from the standard Person correlation, we take into account the influence of node's degree. Substituting this new definition of similarity for the standard Person correlation, we propose a modified collaborative filtering (MCF). Based on a benchmark database, we demonstrate the great improvement of algorithmic accuracy for both user-based MCF and object-based MCF.