Published: May 4, 2025
Language: Английский
Published: May 4, 2025
Language: Английский
Alexandria Engineering Journal, Journal Year: 2025, Volume and Issue: 118, P. 354 - 370
Published: Jan. 23, 2025
Language: Английский
Citations
0ACM transactions on office information systems, Journal Year: 2025, Volume and Issue: unknown
Published: April 29, 2025
Music recommender systems play a critical role in music streaming platforms by providing users with that they are likely to enjoy. Recent studies have shown user emotions can influence users’ preferences for moods. However, existing emotion-aware (EMRSs) explicitly or implicitly assume actual emotional states expressed through identical words homogeneous. They also mood homogeneous under the same state. In this article, we propose four types of heterogeneity an EMRS should account for: emotion across users, within user, preference and user. We further Heterogeneity-aware Deep Bayesian Network (HDBN) model these assumptions. The HDBN mimics user’s decision process choosing components: personalized prior distribution modeling, posterior grouping, neural network-based prediction. constructed two datasets, called EmoMusicLJ EmoMusicLJ-small, validate our method. Extensive experiments demonstrate method significantly outperforms baseline approaches on metrics HR, Precision, NDCG, MRR. Ablation case effectiveness HDBN. source code datasets available at https://github.com/jingrk/HDBN .
Language: Английский
Citations
0Published: May 4, 2025
Language: Английский
Citations
0