Towards the Development of Music Mood Classification of Original Pilipino Music (OPM) Songs Based on Audio and Lyrics Keyword
Authors:
Open Access
- 21 August 2021
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 87-90
Abstract
This paper presents music mood classification of Original Pilipino Music (OPM) songs, particularly Filipino songs using audio and lyrics information. The song's mood is expressed utilizing musical features, but a relevant part also seems to be conveyed by the keywords to its lyrics. The study evaluates with the help of two music teachers and music analysts each factor independently. It explores the possibility of combining both, using Natural Language Processing and Music Information Retrieval techniques. It shows that standard separation-based strategies and Latent Semantic Analysis can group the verses essentially superior to random. Yet, the exhibition is still very substandard compared to that of sound-based systems. The study presents a technique dependent on contrasts between language models that gives performances closer to sound-based classifiers—in addition, interwinding this in a multimodal framework, which is audio and text. It permits an improvement in the general execution. We exhibit that verses and sound data are corresponding and can be joined to improve an ordered framework.Keywords
This publication has 4 references indexed in Scilit:
- Automatic mood detection and tracking of music audio signalsIEEE Transactions on Audio, Speech, and Language Processing, 2005
- Core affect and the psychological construction of emotion.Psychological Review, 2003
- Another stemmerACM SIGIR Forum, 1990
- An algorithm for suffix strippingProgram: electronic library and information systems, 1980