A contrastive news recommendation framework based on curriculum learning

• DOI: 10.1007/s11257-024-09422-0
• Publication Date: 2024-12-28T16:00:01Z
• Authors: Xingran Zhou; Nankai Lin; Weixiong Zheng; Dong Zhou; Aimin Yang

Abstract News recommendation is a form of smart technology designed to offer users news content that matches their interests according to their preferences and interests. Nevertheless, current methodologies exhibit significant limitations. Traditional models often rely on simple random negative sampling for training, an approach that insufficiently captures the patterns and preferences of users' clicking behavior, thereby undermining the model's effectiveness. Furthermore, these systems often face challenges in insufficient modeling due to the limited nature of user interactions. Considering these challenges, this paper presents a contrastive news recommendation framework based on curriculum learning(CNRCL). Specifically, we relate the negative sampling process to users’ interests, and employ curriculum learning to guide the negative sampling procedure. To address the issue of insufficient user interest modeling, we propose to use contrastive learning to bring the user closer to news that is similar to the candidate news, thus enhancing the model's accuracy in predicting user interests, compensating for limited click behaviour. Evaluation of our proposed CNRCL model on the MIND benchmark dataset reveals its superiority over existing news recommendation methods.