In the realm of music recommendation, sequential recommender systems have shown promise in capturing the dynamic nature of music consumption. Nevertheless, traditional Transformer-based models, such as SASRec and BERT4Rec, while effective, encounter challenges due to the unique characteristics of music listening habits. In fact, existing models struggle to create a coherent listening experience due to rapidly evolving preferences. Moreover, music consumption is characterized by a prevalence of repeated listening, i.e. users frequently return to their favourite tracks, an important signal that could be framed as individual or personalized popularity. This paper addresses these challenges by introducing a novel approach that incorporates personalized popularity information into sequential recommendation. By combining user-item popularity scores with model-generated scores, our method effectively balances the exploration of new music with the satisfaction of user preferences. Experimental results demonstrate that a Personalized Most Popular recommender, a method solely based on user-specific popularity, outperforms existing state-of-the-art models. Furthermore, augmenting Transformer-based models with personalized popularity awareness yields superior performance, showing improvements ranging from 25.2% to 69.8%. The code for this paper is available at https://github.com/sisinflab/personalized-popularity-awareness.
In today’s era of information overload, personalized news recommendation systems are crucial for connecting users with relevant content. The dynamic nature of user interests and the fleeting popularity of news articles pose significant challenges to accurate prediction. For this reason, the RecSys 2024 Challenge aims to inspire innovative solutions in this field. This study presents DIVAN (Deep-Interest Virality-Aware Network), our solution for the RecSys 2024 Challenge, combining a Deep Interest Network (DIN) for personalized user interest representation with a Virality-aware Click Predictor that utilizes temporal features to estimate click probability based on news popularity. A user-specific weight balances the influence of DIN and virality-based predictions, enhancing personalization and accuracy. Experiments on the Ekstra Bladet dataset from the Challenge demonstrate how promising DIVAN is in accuracy and beyond-accuracy performance.