Lyric document embeddings for music tagging

2021/11/29 by McVicar, Matt, Di Giorgi, Bruno, Dundar, Baris +1 · 1 citation
#Computation and Language (cs.CL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2112.11436

Abstract

We present an empirical study on embedding the lyrics of a song into a fixed-dimensional feature for the purpose of music tagging. Five methods of computing token-level and four methods of computing document-level representations are trained on an industrial-scale dataset of tens of millions of songs. We compare simple averaging of pretrained embeddings to modern recurrent and attention-based neural architectures. Evaluating on a wide range of tagging tasks such as genre classification, explicit content identification and era detection, we find that averaging word embeddings outperform more complex architectures in many downstream metrics.

Cited by

• Lyrics Matter: Exploiting the Power of Learnt Representations for Music Popularity Prediction

Related

• Musical Word Embedding for Music Tagging and Retrieval
• Automatic tagging using deep convolutional neural networks
• Musical Word Embedding: Bridging the Gap between Listening Contexts and Music
• Perceptual Musical Features for Interpretable Audio Tagging
• Semantic-Aware Interpretable Multimodal Music Auto-Tagging