SABRE: Cross-Domain Crowdsourcing Platform for Recommendation Services

Abstract

Existing recommendation services place emphasis on personalization to achieve promising accuracy of recommendations. This study aims to exploit the user cognition similarity across multiple domains. The purpose is to leverage this information to enhance the user-based collaborative filtering algorithm for cross-domain recommendation services. The main idea of this is i) to collect feedback from users across multiple domains to represent user cognition; ii) to establish a user cognition-based collaborative filtering (UCCF) model for the multi-domain recommendation; iii) generating recommendations in the target domain. The experimental results demonstrate that the prediction performance of the proposed model outperforms in comparison with all baseline methods. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.