Incorporating User-Item Similarity in Hybrid Neighborhood-based Recommendation System

Nghia Duong Tan, Giang Do Truong, Nam Doan Nguyen, Nghia Cao Tuan, Hoang Tran Manh, Minh Nguyen Duc, Hieu Dang Quang

Keywords: Recommendation system, Neighborhood-based, Collaborative filtering, Data mining

Abstract: Modern hybrid recommendation systems require a sufficient amount of data. However, several internet privacy issues make users skeptical about sharing their personal information with online service providers. This work introduces various novel methods utilizing the baseline estimate to learn user interests from their interactions. Subsequently, extracted user feature vectors are implemented to estimate the user-item correlations, providing an additional fine-tuning factor for neighborhood-based collaborative filtering systems. Comprehensive experiments show that utilizing the user-item similarity can boost the accuracy of hybrid neighborhood-based systems by at least $2.11\%$ while minimizing the need for tracking users' digital footprints.

One-sentence Summary: This paper proposes various novel methods to learn user interests from their interactions, which are applied to estimate the user-item correlations to improve neighborhood-based collaborative filtering systems.