Analysis of Public Perception of Brazilian Economic Ministers on YouTube Using NLP

• Daniel M. Prates UFOP
• Guilherme O. Santos UFOP
• Vinícius da F. Vieira UFSJ
• Carlos H. G. Ferreira UFOP

DOI: https://doi.org/10.5753/brasnam.2025.8940

Abstract

This study investigates how Brazil’s economic ministers are perceived by the public through comments posted on YouTube. Despite the growing use of social media for political and economic analysis, there is still a gap in the national literature regarding the perception of key economic figures outside electoral contexts. To address this gap, we analyze over 679,000 comments related to Paulo Guedes and Fernando Haddad using Natural Language Processing techniques and a fine-tuned BERT-based model for stance detection. The results reveal distinct patterns of support and criticism for each minister, evidence of polarization in user interactions, and notable temporal variations. The findings highlight YouTube’s potential as a complementary source for understanding public discourse around economic policy in Brazil.

References

Antenucci, D., Cafarella, M., Levenstein, M., Ré, C., and Shapiro, M. D. (2014). Using social media to measure labor market flows. Technical report, NBER.

Baldwin, T., Cook, P., Lui, M., MacKinlay, A., and Wang, L. (2013). How noisy social media text, how diffrnt social media sources? In Mitkov, R. and Park, J. C., editors, International Joint Conference on Natural Language Processing, pages 356–364.

Born, B., Dalal, H., Lamersdorf, N., and Steffen, S. (2023). Monetary policy in the age of social media: A twitter-based inflation analysis. Available at SSRN.

Carter, D., Westcott, M., Boitel, R., and Wang, A. H.-E. (2023). Puppet anchors and china’s youtube information operation. Taiwan Politics.

Celli, F., Stepanov, E., Poesio, M., and Riccardi, G. (2016). Predicting brexit: Classifying agreement is better than sentiment and pollsters. In Proceedings of the Workshop on Computational Modeling of People’s Opinions, Personality, and Emotions in Social Media (PEOPLES).

Chandra, R. and Saini, R. (2021). Biden vs trump: modeling us general elections using bert language model. IEEE access, 9:128494–128505.

Christhie, W., Reis, J. C., Moro, F. B. M. M., and Almeida, V. (2018). Detecção de posicionamento em tweets sobre política no contexto brasileiro. In Brazilian Workshop on Social Network Analysis and Mining (BraSNAM). SBC.

Cinelli, M., De Francisci Morales, G., Galeazzi, A., Quattrociocchi, W., and Starnini, M. (2021). The echo chamber effect on social media. Proceedings of the national academy of sciences.

Cohen, I., Huang, Y., Chen, J., Benesty, J., Benesty, J., Chen, J., Huang, Y., and Cohen, I. (2009). Pearson correlation coefficient. Noise reduction in speech processing, pages 1–4.

Costa, J., Oliveira, G., Fonseca, G., Reis, D., Teixeira, G., Cunha, W., Rocha, L., and Ferreira, C. H. (2025). Characterizing youtube’s role in online gambling promotion: A case study of fortune tiger in brazil. In Proceedings of the 17th ACM Web Science Conference 2025.

da Rosa Jr, J. M., Linhares, R. S., Ferreira, C. H. G., Nobre, G. P., Murai, F., and Almeida, J. M. (2022). Uncovering discussion groups on claims of election fraud from twitter. In International conference on social informatics, pages 320–336. Springer.

Dias, A., Tanure, R. R., Almeida, J. M., Lima, H. C., and Ferreira, C. H. (2024). Análise da percepção do uso de cigarros eletrônicos no brasil por meio de comentários no youtube. In Brazilian Symposium on Multimedia and the Web (WebMedia).

Durrani, N., Sajjad, H., and Dalvi, F. (2021). How transfer learning impacts linguistic knowledge in deep NLP models?

Effing, R., Van Hillegersberg, J., and Huibers, T. (2011). Social media and political participation: are facebook, twitter and youtube democratizing our political systems? In Electronic Participation: Third IFIP WG 8.5 International Conference, ePart 2011, Delft, The Netherlands, August 29–September 1, 2011. Proceedings 3, pages 25–35. Springer.

Eyal, M., Amrami, A., Taub-Tabib, H., and Goldberg, Y. (2021). Bootstrapping relation extractors using syntactic search by examples. arXiv preprint arXiv:2102.05007.

Feitosa, M., Ferreira, C., Gonçalves, G., and Almeida, J. (2022). Análise da percepção das pessoas no twitter sobre ações policiais. In Anais do XI Brazilian Workshop on Social Network Analysis and Mining, pages 73–84, Porto Alegre, RS, Brasil. SBC.

Fleiss, J. L., Levin, B., and Paik, M. C. (2013). Statistical methods for rates and proportions. john wiley & sons.

Franziska B. Keller, David Schoch, S. S. and Yang, J. (2020). Political astroturfing on twitter: How to coordinate a disinformation campaign. Political Communication.

Gorodnichenko, Y., Pham, T., and Talavera, O. (2024). Central bank communication on social media: What, to whom, and how? Journal of Econometrics, page 105869.

Grimminger, L. and Klinger, R. (2021). Hate towards the political opponent: A twitter corpus study of the 2020 us elections on the basis of offensive speech and stance detection. In Proceedings of the Eleventh Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis.

Hossain, T., Logan IV, R. L., Ugarte, A., Matsubara, Y., Young, S., and Singh, S. (2020). COVIDLies: Detecting COVID-19 misinformation on social media. In Proceedings of the 1st Workshop on NLP for COVID-19 (Part 2) at EMNLP 2020.

Hui, L. and Belkin, M. (2021). Evaluation of neural architectures trained with square loss vs cross-entropy in classification tasks. In International Conference on Learning Representations.

Kawintiranon, K. and Singh, L. (2021). Knowledge enhanced masked language model for stance detection. In Proceedings of the 2021 conference of the north american chapter of the association for computational linguistics: human language technologies, pages 4725–4735.

Keet, C. M. (2022). Bootstrapping nlp tools across low-resourced african languages: an overview and prospects. arXiv preprint arXiv:2210.12027.

Krihova, Z. (2013). Media, power, and politics in the digital age. the 2009 presidential election uprising in iran. CyberOrient, 7(1):119–123.

Lai, M., Patti, V., Ruffo, G., and Rosso, P. (2020). brexit: Leave or remain? the role of user’s community and diachronic evolution on stance detection. Journal of Intelligent & Fuzzy Systems.

Landis, J. R. and Koch, G. G. (1977). An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics.

Liu, J., Duan, X., Zhang, R., Sun, Y., Guan, L., and Lin, B. (2021). Relation classification via bert with piecewise convolution and focal loss. Plos one, 16(9):e0257092.

Malagoli, L., Stancioli, J., Ferreira, C. H., Vasconcelos, M., da Silva, A. P. C., and Almeida, J. (2021). Caracterizaçao do debate no twitter sobre a vacinaçao contra a covid-19 no brasil. In Brazilian Workshop on Social Network Analysis and Mining (BraSNAM), pages 55–66. SBC.

Mehta, P. and Pandya, S. (2020). A review on sentiment analysis methodologies, practices and applications. International Journal of Scientific and Technology Research, 9(2):601–609.

Reis, R., Zanetti, D., and Frizzera, L. (2019). Algorítmos e desinformação: o papel do youtube no cenário político brasileiro. In Assoc. Brasileira de Pesquisadores em Comunicação e Política.

Rocha, S., Silva, C., Ferreira, C., Gonçalves, G., and Almeida, J. (2024). Monitorando a opinião pública sobre operações policiais no brasil via comentários de vídeos no youtube. In Anais do XIII Brazilian Workshop on Social Network Analysis and Mining.

Santos, P. D. and Goya, D. H. (2022). Detecção de posicionamento e rotulação automática de usuários do twitter: estudo sobre o embate científico-político no contexto da cpi da covid-19. In Brazilian Workshop on Social Network Analysis and Mining (BraSNAM), pages 49–60. SBC.

Shen, J., Najand, M., Dong, F., and He, W. (2017). News and social media emotions in the commodity market. Review of Behavioral Finance, 9(2):148–168.

Souza, F., Nogueira, R., and Lotufo, R. (2020). Bertimbau: Pretrained bert models for brazilian portuguese. In Intelligent Systems.