A musical mood trajectory estimation method using lyrics and acoustic features

Naoki Nishikawa, Katsutoshi Itoyama, Hiromasa Fujihara, Masataka Goto, Tetsuya Ogata, Hiroshi G. Okuno

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

In this paper, we present a new method that represents an overall musical time-varying impression of a song by a pair of mood trajectories estimated from lyrics and audio signals. The mood trajectory of the lyrics is obtained by using the probabilistic latent semantic analysis (PLSA) to estimate topics (representing impressions) from words in the lyrics. The mood trajectory of the audio signals is estimated from acoustic features by using the multiple linear regression analysis. In our experiments, the mood trajectories of 100 songs in Last.fm's Best of 2010 were estimated. The detailed analysis of the 100 songs confirms that acoustic features provide more accurate mood trajectory and the 21% resulting mood trajectories are matched to realistic musical mood available at Last.fm.

Original languageEnglish
Title of host publicationMM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11
Pages51-56
Number of pages6
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event2011 ACM Multimedia Conference, MM'11 and Co-Located Workshops - 1st International ACM Workshop on Music Information Retrieval with User-Centered and Multimodal Strategies, MIRUM'11 - Scottsdale, AZ
Duration: 2011 Nov 282011 Dec 1

Other

Other2011 ACM Multimedia Conference, MM'11 and Co-Located Workshops - 1st International ACM Workshop on Music Information Retrieval with User-Centered and Multimodal Strategies, MIRUM'11
CityScottsdale, AZ
Period11/11/2811/12/1

Fingerprint

Acoustics
Trajectories
Linear regression
Regression analysis
Semantics
Experiments

Keywords

  • Lyrics and audio signals music information retrieval
  • Mood trajectory
  • Musical mood estimation
  • Musical mood representation
  • Time-varying impression

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Human-Computer Interaction

Cite this

Nishikawa, N., Itoyama, K., Fujihara, H., Goto, M., Ogata, T., & Okuno, H. G. (2011). A musical mood trajectory estimation method using lyrics and acoustic features. In MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11 (pp. 51-56) https://doi.org/10.1145/2072529.2072543

A musical mood trajectory estimation method using lyrics and acoustic features. / Nishikawa, Naoki; Itoyama, Katsutoshi; Fujihara, Hiromasa; Goto, Masataka; Ogata, Tetsuya; Okuno, Hiroshi G.

MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11. 2011. p. 51-56.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nishikawa, N, Itoyama, K, Fujihara, H, Goto, M, Ogata, T & Okuno, HG 2011, A musical mood trajectory estimation method using lyrics and acoustic features. in MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11. pp. 51-56, 2011 ACM Multimedia Conference, MM'11 and Co-Located Workshops - 1st International ACM Workshop on Music Information Retrieval with User-Centered and Multimodal Strategies, MIRUM'11, Scottsdale, AZ, 11/11/28. https://doi.org/10.1145/2072529.2072543
Nishikawa N, Itoyama K, Fujihara H, Goto M, Ogata T, Okuno HG. A musical mood trajectory estimation method using lyrics and acoustic features. In MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11. 2011. p. 51-56 https://doi.org/10.1145/2072529.2072543
Nishikawa, Naoki ; Itoyama, Katsutoshi ; Fujihara, Hiromasa ; Goto, Masataka ; Ogata, Tetsuya ; Okuno, Hiroshi G. / A musical mood trajectory estimation method using lyrics and acoustic features. MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11. 2011. pp. 51-56
@inproceedings{797e8f0fae884997854020b88b08bfda,
title = "A musical mood trajectory estimation method using lyrics and acoustic features",
abstract = "In this paper, we present a new method that represents an overall musical time-varying impression of a song by a pair of mood trajectories estimated from lyrics and audio signals. The mood trajectory of the lyrics is obtained by using the probabilistic latent semantic analysis (PLSA) to estimate topics (representing impressions) from words in the lyrics. The mood trajectory of the audio signals is estimated from acoustic features by using the multiple linear regression analysis. In our experiments, the mood trajectories of 100 songs in Last.fm's Best of 2010 were estimated. The detailed analysis of the 100 songs confirms that acoustic features provide more accurate mood trajectory and the 21{\%} resulting mood trajectories are matched to realistic musical mood available at Last.fm.",
keywords = "Lyrics and audio signals music information retrieval, Mood trajectory, Musical mood estimation, Musical mood representation, Time-varying impression",
author = "Naoki Nishikawa and Katsutoshi Itoyama and Hiromasa Fujihara and Masataka Goto and Tetsuya Ogata and Okuno, {Hiroshi G.}",
year = "2011",
doi = "10.1145/2072529.2072543",
language = "English",
isbn = "9781450309868",
pages = "51--56",
booktitle = "MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11",

}

TY - GEN

T1 - A musical mood trajectory estimation method using lyrics and acoustic features

AU - Nishikawa, Naoki

AU - Itoyama, Katsutoshi

AU - Fujihara, Hiromasa

AU - Goto, Masataka

AU - Ogata, Tetsuya

AU - Okuno, Hiroshi G.

PY - 2011

Y1 - 2011

N2 - In this paper, we present a new method that represents an overall musical time-varying impression of a song by a pair of mood trajectories estimated from lyrics and audio signals. The mood trajectory of the lyrics is obtained by using the probabilistic latent semantic analysis (PLSA) to estimate topics (representing impressions) from words in the lyrics. The mood trajectory of the audio signals is estimated from acoustic features by using the multiple linear regression analysis. In our experiments, the mood trajectories of 100 songs in Last.fm's Best of 2010 were estimated. The detailed analysis of the 100 songs confirms that acoustic features provide more accurate mood trajectory and the 21% resulting mood trajectories are matched to realistic musical mood available at Last.fm.

AB - In this paper, we present a new method that represents an overall musical time-varying impression of a song by a pair of mood trajectories estimated from lyrics and audio signals. The mood trajectory of the lyrics is obtained by using the probabilistic latent semantic analysis (PLSA) to estimate topics (representing impressions) from words in the lyrics. The mood trajectory of the audio signals is estimated from acoustic features by using the multiple linear regression analysis. In our experiments, the mood trajectories of 100 songs in Last.fm's Best of 2010 were estimated. The detailed analysis of the 100 songs confirms that acoustic features provide more accurate mood trajectory and the 21% resulting mood trajectories are matched to realistic musical mood available at Last.fm.

KW - Lyrics and audio signals music information retrieval

KW - Mood trajectory

KW - Musical mood estimation

KW - Musical mood representation

KW - Time-varying impression

UR - http://www.scopus.com/inward/record.url?scp=84555188101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84555188101&partnerID=8YFLogxK

U2 - 10.1145/2072529.2072543

DO - 10.1145/2072529.2072543

M3 - Conference contribution

SN - 9781450309868

SP - 51

EP - 56

BT - MM'11 - Proceedings of the 2011 ACM Multimedia Conference and Co-Located Workshops - MIRUM 2011 Workshop, MIRUM'11

ER -