User-Item Correlation in Hybrid Neighborhood-Based Recommendation System with Synthetic User Data

Abstract

Recommendation systems have been widely adopted to help users with the information overload from the large volume of online multimedia content by providing them with appropriate options. While modern hybrid recommendation systems require an immense amount of data, several existing online privacy issues make users skeptical about sharing their personal information with service providers. This work introduces various novel methods utilizing the baseline estimate to learn user interests from their interactions. Subsequently, synthetic 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.