MUSIC SENTIMENT ANALYSIS
DOI:
https://doi.org/10.24867/27BE20BlagojevicKeywords:
Sentiment analysis, RNN, Word2Vec, GloVe, LLM, GPTAbstract
This paper describes a comparison of approaches for classifying the sentiment of songs. For the purposes of the study, a dataset was created that includes audio characteristics, lyrics and song sentiments. Classic NLP approaches for text classification and a generative approach using GPT models were compared.
References
[1] E. Çano, “Text-based Sentiment Analysis and Music Emotion Recognition,” arXiv:1810.03031 [cs], Jun. 2018, doi: https://doi.org/10.6092/polito/porto/2709436.
[2] M. McVicar, B. Di Giorgi, B. Dundar, and M. Mauch, “Lyric document embeddings for music tagging,” arXiv.org, Nov. 29, 2021. https://arxiv.org/abs/2112.11436 (accessed Nov. 07, 2023).
[2] M. McVicar, B. Di Giorgi, B. Dundar, and M. Mauch, “Lyric document embeddings for music tagging,” arXiv.org, Nov. 29, 2021. https://arxiv.org/abs/2112.11436 (accessed Nov. 07, 2023).
Downloads
Published
2024-06-06
Issue
Section
Electrotechnical and Computer Engineering
