AMAZON ON TELEGRAM
A BERT-BASED TOPIC MODELING DASHBOARD
Keywords:
social media, artificial intelligence, natural language processing, AmazonAbstract
The Amazon occupies a central position in contemporary disputes over climate, sovereignty, conservation, and development, which makes it fertile ground for disinformation narratives. This study presents a methodological strategy for exploring large volumes of text, applied to messages about the Amazon posted by public Telegram channels in a multilingual corpus (Portuguese, Spanish, and English). Building on a previous study, the research advances the application of BERT-based topic modeling, coupled with the construction of an interactive dashboard. The aim is to discuss how semantic representations, thematic clusters, and exploratory visualizations can support the interpretation of extensive and heterogeneous textual data. The dashboard makes it possible to observe topic evolution over time, hierarchical structures, inter-topic distances, and distribution across channels. We argue that combining machine learning, visualization, and qualitative analysis contributes to research in Information Science and climate disinformation, taking the Amazon as an empirical case study and offering a replicable workflow for other thematic corpora.
References
BLEI, David M. Probabilistic topic models. Communications of the ACM, New York, v. 55, n. 4, p. 77-84, abr. 2012. Disponível em: http://dl.acm.org/citation.cfm?doid=2133806.2133826. Acesso em: 12 dez. 2018.
BLEI, David M.; NG, Andrew Y.; JORDAN, Michael I. Latent Dirichlet allocation. Journal of Machine Learning Research, Cambridge, v. 3, p. 993-1022, mar. 2003. Disponível em: https://www.jmlr.org/papers/v3/blei03a.html. Acesso em: 17 mar. 2026.
BONALDO, Manoela; QUEIROZ, Gustavo; BORBA, Bibiana. Bruno Pereira e Dom Phillips: o que se sabe sobre o assassinato do jornalista e do indigenista. Estadão, São Paulo, 23 jun. 2022. Disponível em: https://www.estadao.com.br/politica/bruno-pereira-e-dom-phillips-o-que-se-sabe-sobre-o-assassinato-do-jornalista-e-do-indigenista/. Acesso em: 23 out. 2025.
BOYD, danah; CRAWFORD, Kate. Critical questions for big data. Information, Communication & Society, Abingdon, v. 15, n. 5, p. 662-679, jun. 2012. Disponível em: https://www.tandfonline.com/doi/full/10.1080/1369118X.2012.678878. Acesso em: 28 set. 2021.
CARNEIRO, Bia; TUCCI, Giulia. The Amazon biome in messages: public policy and communication frames on Telegram. Rome: Bioversity International; International Center for Tropical Agriculture (CIAT), set. 2025. 28 p. Working paper. Disponível em: https://hdl.handle.net/10568/177294. Acesso em: 16 jun. 2026.
DAVIS, Charles; YRA, Goia; SILVA, Caio; GUIMARÃES, Mariana et al. Social media are fuelling the Amazon’s destruction. Nature, London, v. 580, n. 7803, p. 321, abr. 2020. Disponível em: https://www.nature.com/articles/d41586-020-01078-1. Acesso em: 25 ago. 2025.
DEVLIN, Jacob; CHANG, Ming-Wei; LEE, Kenton; TOUTANOVA, Kristina. BERT: pre-training of deep bidirectional transformers for language understanding. In: CONFERENCE OF THE NORTH AMERICAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS: HUMAN LANGUAGE TECHNOLOGIES (NAACL-HLT), 2019, Minneapolis. Anais [...]. Minneapolis: Association for Computational Linguistics, 2019. v. 1 (long and short papers), p. 4171-4186. Disponível em: https://aclanthology.org/N19-1423/. Acesso em: 15 jun. 2026.
FERREIRA, Adison. Desinformação sobre a Amazônia: relatório do coletivo Intervozes mostra como fake news se espalham na região. InfoAmazonia, Amazônia, 4 maio 2023. Disponível em: http://infoamazonia.org/podcast/relatorio-mostra-como-desinformacao-se-espalha-na-amazonia/. Acesso em: 25 ago. 2025.
FRANCO, Marco A.; RIZZO, Luciana V.; TEIXEIRA, Márcio J.; ARTAXO, Paulo et al. How climate change and deforestation interact in the transformation of the Amazon rainforest. Nature Communications, London, v. 16, n. 1, p. 7944, set. 2025. Disponível em: https://www.nature.com/articles/s41467-025-63156-0. Acesso em: 27 abr. 2026.
GROOTENDORST, Maarten. BERTopic. [S. l.]: GitHub, 2022. Software. Disponível em: https://github.com/MaartenGr/BERTopic. Acesso em: 13 maio 2023.
MURSHED, Belal Abdullah Hezam; MALLAPPA, Suresha; ABAWAJY, Jemal; SAIF, Mufeed Ahmed Naji et al. Short text topic modelling approaches in the context of big data: taxonomy, survey, and analysis. Artificial Intelligence Review, Dordrecht, v. 56, n. 6, p. 5133-5260, jun. 2023. Disponível em: https://link.springer.com/article/10.1007/s10462-022-10254-w. Acesso em: 4 maio 2026.
PYTHON SOFTWARE FOUNDATION. Python.org. Beaverton: PSF, [20--]. Disponível em: https://www.python.org/. Acesso em: 3 jan. 2024.
ROGERS, Richard. Deplatforming: following extreme Internet celebrities to Telegram and alternative social media. European Journal of Communication, London, v. 35, n. 3, p. 213-229, jun. 2020. Disponível em: http://journals.sagepub.com/doi/10.1177/0267323120922066. Acesso em: 1 jul. 2021.
ROGERS, Richard. Digital methods. Cambridge: MIT Press, 2013. 280 p.
RUSCICA, Giosuè; TUCCI, Giulia; CARNEIRO, Bia. TeleCatch: an open-access software for visualizing, filtering and extracting Telegram messages data. Software Impacts, Amsterdam, v. 23, p. 100736, mar. 2025. Disponível em: https://www.sciencedirect.com/science/article/pii/S2665963824001246. Acesso em: 7 jan. 2025.
SILVA-JUNIOR, Celso H. L.; ILVA, Fabrício B.; ARISI, Barbara Maisonnave; MATAVELI, Guilherme et al. Brazilian Amazon indigenous territories under deforestation pressure. Scientific Reports, London, v. 13, p. 5851, abr. 2023. Disponível em: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10085996/. Acesso em: 25 ago. 2025.
SOUZA, Marcos de; SOUZA, Renato Rocha. Modelagem de tópicos: resumir e organizar corpus de dados por meio de algoritmos de aprendizagem de máquina. Múltiplos Olhares em Ciência da Informação, Belo Horizonte, v. 9, n. 2, 2019. Disponível em: https://periodicos.ufmg.br/index.php/moci/article/view/19138. Acesso em: 11 jun. 2026.
TREEN, Kathie; WILLIAMS, Hywel T. P.; O’NEILL, Saffron J. Online misinformation about climate change. WIREs Climate Change, Hoboken, v. 11, n. 5, p. e665, 2020. Disponível em: https://onlinelibrary.wiley.com/doi/abs/10.1002/wcc.665. Acesso em: 7 abr. 2025.
TUCCI, Giulia. Telegram channels data. [S. I.]: OSF, 2025. Conjunto de dados. Disponível em: https://osf.io/ys4j9/. Acesso em: 26 ago. 2025.
TUCCI, Giulia; GOUVEIA, Fábio Castro. O discurso do bolsonarismo nas eleições 2022: uma investigação da desinformação viral em grupos de Telegram. Em Questão, Porto Alegre, v. 31, 29 abr. 2025. Disponível em: https://seer.ufrgs.br/index.php/EmQuestao/article/view/141754. Acesso em: 12 ago. 2025.