Automatic Playlist Generation: A Comprehensive Crux of a Methodology for Technology Trend

초록

The rapid expansion of large-scale music data on streaming platforms has significantly fueled the development of music recommendation systems, which aim to enhance music exploration by accurately reflecting user preferences. A central challenge in this domain is improving the accuracy of Automatic Playlist Generation, which involves generating new playlists that align with user preferences while maintaining the musical characteristics of previously listened tracks. Various Automatic Playlist Generation methods have been proposed, utilizing different data types and algorithms. This paper provides an classification of existing Automatic Playlist Generation approaches based on neural network architectures, model principles, characteristics, and applications. Additionally, it discusses the evaluation methodologies used to assess Automatic Playlist Generation performance and compare the quality of generated playlists. The paper also explores real-world applications of Automatic Playlist Generation and the potential for integrating these systems across different platforms and devices, emphasizing the importance of addressing the remaining challenges in future research. © 2025 IEEE.

키워드