从分类到推荐：音频嵌入模型在基于内容的音乐推荐中的实证分析
原标题：From Classification to Recommendation: Empirical Analysis of Audio Embedding Models Application for Content-Based Music Recommendation
AI 摘要
该研究系统评估了六种预训练音频编码器在三种音乐推荐系统（基于内容、序列和基于语义ID的生成式推荐系统）中的有效性。研究发现，音频-文本对齐和音乐领域的表示在直接使用预训练嵌入几何时通常更有效，而基于交互的序列训练会显著缩小编码器间的性能差异。此外，增加语义ID容量并不总能提升生成式推荐系统性能，甚至可能引入不稳定性。这些发现为现代音乐推荐系统选择音频编码器和设计音频派生语义ID提供了实用指导。
正文节选
Computer Science > Information Retrieval Title:From Classification to Recommendation: Empirical Analysis of Audio Embedding Models Application for Content-Based Music Recommendation View PDF HTML (experimental) Abstract:Pretrained audio representation models learned from large-scale corpora have achieved strong performance in audio classification and understanding. However, most existing models are optimized for objectives such as masked prediction, contrastive learning, or audio-tex