Sector-Specific Financial Sentiment Lexicons for Portuguese: Evidence from Brazilian Stock News

• Antônio Artur De Souza UFMG
• Mateus Binda UFMG
• Adriana S. Pagano UFMG

DOI: https://doi.org/10.5753/stil.2026.25274

Resumo

This paper investigates whether sector-specific financial sentiment lexicons improve the analysis of Brazilian stock-market news in Portuguese. We use 19,460 news items about 38 firms in the financial, electric, and manufacturing sectors, published between 2020 and 2025. Starting from OpLexicon, we build sector-specific lexicons through unsupervised expansion based on PMI and conditional word-polarity association. The lexicons are evaluated by lexical heterogeneity and through LSTM price forecasting. Results show sectoral variation and moderate classification performance, but limited predictive gains. Sentiment improves forecasting only in manufacturing, while other sectors do not benefit from its inclusion. The findings indicate that sector-specific lexicons are useful resources, although their predictive value are context dependent.

Palavras-chave: Sentiment analysis, financial lexicons, Portuguese NLP, domain adaptation, stock-market news

Referências

B3 (2025). Critério de classificação.

de Melo, T. (2021). Sentiprodbr: Building domain-specific sentiment lexicons for the portuguese language. In Anais do Simpósio Brasileiro de Banco de Dados, pages 349–354.

de Melo, T. (2022). Sentilexbr: An automatic methodology of building sentiment lexicons for the portuguese language. Journal of Information and Data Management, 13(3):439–449.

de Sousa, T. M. and Fernandes, D. S. A. (2023). Expansão automática de léxico para análise de sentimentos de twitter no domínio do mercado financeiro brasileiro. In Anais da Escola Regional de Informática de Goiás.

Deveikyte, J., Geman, H., Piccari, C., and Provetti, A. (2022). A sentiment analysis approach to the prediction of market volatility. Frontiers in Artificial Intelligence, 5:1–10.

Gao, Y., Zhao, C., Sun, B., and Zhao, W. (2022). Effects of investor sentiment on stock volatility: new evidences from multi-source data in china’s green stock markets. Financial Innovation, 8(1):77.

Hsu, Y.-J., Lu, Y.-C., and Yang, J. J. (2021). News sentiment and stock market volatility. Review of Quantitative Finance and Accounting, 57(3):1093–1122.

Huang, J.-Y. and Liu, J.-H. (2020). Using social media mining technology to improve stock price forecast accuracy. Journal of Forecasting, 39(1):104–116.

Janková, Z. (2023). Critical review of text mining and sentiment analysis for stock market prediction. Journal of Business Economics and Management, 24(1):177–198.

Januário, B. A. et al. (2022). Sentiment analysis applied to news from the brazilian stock market. IEEE Latin America Transactions, 20(3):512–518.

Kos, S. R., Scarpin, J. E., and Pinto, J. S. d. P. (2019). Proposta de um dicionário de termos de emoção em textos financeiros. Revista de Ciências da Administração, 21(55):147–164.

Loughran, T. and McDonald, B. (2011). When is a liability not a liability? textual analysis, dictionaries, and 10-ks. Journal of Finance, 66(1):35–65.

Peres, V., Vieira, R., and Bordini, R. (2019). Análises de sentimentos: abordagem lexical de classificação de opinião no contexto mercado financeiro brasileiro. Anais do Workshop of Artificial Intelligence Applied to Finance.

Ranco, G. et al. (2015). The effects of twitter sentiment on stock price returns. PLoS ONE, 10(9):e0138441.

Shapiro, A. H., Sudhof, M., and Wilson, D. J. (2022). Measuring news sentiment. Journal of Econometrics, 228(2):221–243.

Souza, M. et al. (2011). Construction of a portuguese opinion lexicon from multiple resources. In Proceedings of the Brazilian Symposium in Information and Human Language Technology.

Zhao, C., Kang, L., Xi, X., Du, S., and Li, J. (2025). Investor sentiment and stock market volatility: exploring the relationship using sentiment analysis of stock bar comments. Finance Research Open, 1:100016.