Predicting Sentiments in Spotify Comments: A Comparative Analysis of Machine Learning Models
Year: 2024
Type: article
Abstract: Using data from user sentences on Spotify, this work explores through Natural Language Processing positive and negative sentiments in each comment. We compare different statistical modeling and Machine Learning techniques, identifying the ones with the greatest accuracy in predicting sentiments. As a result, the assessment supports most of the sentences presented with negative connotations. As for modeling, the Logistic Regression and Random Forest models resulted in better accuracy.
Source: Socioeconomic Analytics
Authors: Filipe Augusto Felix de Queiroz, Igor Barbosa Negreiros, Giovana de Souza, Débora Cordeiro de Sousa, Sílvio Fernando Alves Xavier Júnior
Institution: Universidade Estadual da Paraíba
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4405849100