We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.IR

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Transformer-based approach towards music emotion recognition from lyrics

Abstract: The task of identifying emotions from a given music track has been an active pursuit in the Music Information Retrieval (MIR) community for years. Music emotion recognition has typically relied on acoustic features, social tags, and other metadata to identify and classify music emotions. The role of lyrics in music emotion recognition remains under-appreciated in spite of several studies reporting superior performance of music emotion classifiers based on features extracted from lyrics. In this study, we use the transformer-based approach model using XLNet as the base architecture which, till date, has not been used to identify emotional connotations of music based on lyrics. Our proposed approach outperforms existing methods for multiple datasets. We used a robust methodology to enhance web-crawlers' accuracy for extracting lyrics. This study has important implications in improving applications involved in playlist generation of music based on emotions in addition to improving music recommendation systems.
Comments: Appearing in Proceedings of the 43rd European Conference On Information Retrieval (ECIR) 2021
Subjects: Information Retrieval (cs.IR)
Journal reference: Lecture Notes in Computer Science, 12657 (2021) 167-175
DOI: 10.1007/978-3-030-72240-1_12
Cite as: arXiv:2101.02051 [cs.IR]
  (or arXiv:2101.02051v1 [cs.IR] for this version)

Submission history

From: Yudhik Agrawal [view email]
[v1] Wed, 6 Jan 2021 14:07:45 GMT (431kb,D)

Link back to: arXiv, form interface, contact.