Sentiment Evaluation of Applications: A Comparison of Large Language Models

• Kalidsa B. de Oliveira UFSM
• Gabriel M. Lunardi UFSM
• Williamson Silva UNIPAMPA

DOI: https://doi.org/10.5753/erbd.2025.7011

Abstract

The expansion of e-commerce in Brazil has driven the demand for more efficient strategies to understand consumer perspectives. In this context, sentiment analysis emerges as an essential tool for examining user opinions on products and services. This research evaluates the performance of different Large Language Models (LLMs) with low computational cost, in the task of sentiment analysis.

References

Bergmann, D. (2024). O que é ajuste fino? [link]. (Accessed on 08/02/2025).

Mendes, K. C. P., Paucar, V. L., and Filho, R. N. D. C. (2022). Análise de sentimentos de tweets utilizando diferentes técnicas de deep learning. In Anais da Escola Regional de Engenharia de Software (ERES).

Siqueira, V., Costa, R. H., Soares, T., Lunardi, G., and Silva, W. (2024a). Dataset anotado de sentimentos a partir de comentários de aplicativos móveis. In Anais do VI Dataset Showcase Workshop, pages 65–76, Porto Alegre, RS, Brasil. SBC.

Siqueira, V., Lunardi, G. M., Silva, W., Hentges Costa, R. L., and Tales, S. S. (2024b). A dataset of polarities and emotions from brazilian portuguese play store reviews. DOI: 10.5281/zenodo.10823148. (Accessed on 13/02/2025.).

Souto Moreira, L., Machado Lunardi, G., de Oliveira Ribeiro, M., Silva, W., and Paulo Basso, F. (2023). A study of algorithm-based detection of fake news in brazilian election: Is bert the best. IEEE Latin America Transactions, 21(8):897–903.

Átilas Barros (2024). Desvendando narrativas e representações: a ia aplicada à análise de dados qualitativos. Educare et Sabere, 2(2). (Accessed on 08/02/2025).