Algorithmic analysis of YouTube music comments: measurement and applications

Stephane Gauvin

Economia della Cultura, 2025, issue 1, 81-90

Abstract: Algorithmic sentiment analysis, which automatically detects the emotional tone of textual data, is an important tool for understanding users' opinions. Recent advances in machine learning have greatly improved model accuracy. We show that a carefully trained BERT model outperforms Generative Pretrained Transformers (GPTs) - and humans - in inferring the sentiment of comments left on YouTube music videos, with an almost perfect FI-score of 0.99. We apply this model to a corpus of 700 million English-language comments left on YouTube's Official Artist Channels, showing that inferences are valid and that, counter-intuitively, sentiment towards superstars is lower than the global average.

Keywords: transformers; BERT; sentiment; music; YouTube; comments (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.rivisteweb.it/download/article/10.1446/118111 (application/pdf)
https://www.rivisteweb.it/doi/10.1446/118111 (text/html)
Access to full text is restricted to subscribers

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:mul:jkrece:doi:10.1446/118111:y:2025:i:1:p:81-90

Access Statistics for this article

Economia della Cultura is currently edited by Paolo Leon

More articles in Economia della Cultura from Società editrice il Mulino
Bibliographic data for series maintained by ().