Information filtering in sparse online systems: recommendation via semi-local diffusion

TL;DR

This paper introduces a semi-local diffusion recommendation algorithm designed to address data sparsity in online systems, demonstrating significant improvements over existing methods especially for users with few interactions.

Contribution

The paper proposes a novel semi-local diffusion algorithm for sparse bipartite networks and introduces two personalized variants that enhance recommendation accuracy.

Findings

Outperforms state-of-the-art methods on sparse datasets

Significantly improves recommendations for small-degree users

Highlights differences between sparse and dense online systems

Abstract

With the rapid growth of the Internet and overwhelming amount of information and choices that people are confronted with, recommender systems have been developed to effectively support users' decision-making process in the online systems. However, many recommendation algorithms suffer from the data sparsity problem, i.e. the user-object bipartite networks are so sparse that algorithms cannot accurately recommend objects for users. This data sparsity problem makes many well-known recommendation algorithms perform poorly. To solve the problem, we propose a recommendation algorithm based on the semi-local diffusion process on a user-object bipartite network. The numerical simulation on two sparse datasets, Amazon and Bookcross, show that our method significantly outperforms the state-of-the-art methods especially for those small-degree users. Two personalized semi-local diffusion methods are proposed which further improve the recommendation accuracy. Finally, our work indicates that sparse online systems are essentially different from the dense online systems, all the algorithms and conclusions based on dense data should be rechecked again in sparse data.