Análise Comparativa Multiplataforma do Debate Sobre Adultização Infantil: Um Estudo de Caso de Vídeo no YouTube, TikTok e Instagram

• Henrique Alves Campos UFV
• João Francisco Hecksher Olivetti UFV
• Philipe de Freitas Melo UFV

DOI: https://doi.org/10.5753/webmedia.2026.30899

Resumo

This paper presents a longitudinal case study on public engagement around the topic of child adultization in Brazil, analyzing 32,791 comments collected on YouTube, TikTok, and Instagram in response to a video published simultaneously by content creator Felca in August 2025. The fact that the content was published simultaneously on all three platforms allows for a rich and controlled comparison of each social network’s discursive patterns, eliminating the video as a confounding variable. We developed a cross-platform collection and analysis pipeline integrating data scraping, sentiment analysis using the pysentimiento model, and topic modeling with BERTopic. The modeling identified 19 interpretable topics, predominantly expressing emotional support for the creator, with substantive debate representing a smaller share of the corpus. Crucially, results revealed that each platform’s structural characteristics were strongly associated with distinct engagement patterns: comments on Instagram were highly reactive and positive; on TikTok, discussion concentrated on specific niche cases; and YouTube emerged as the primary space for deeper textual analysis and substantive debate. These findings highlight the methodological value of crossplatform analysis, demonstrating how single-network studies may yield partial views biased by platform architecture.

Palavras-chave: Child Adultization, Sentiment Analysis, Topic Modeling, Social Media, BERTopic, Cross-Platform Analysis

Referências

Despoina Antonakaki and Sotiris Ioannidis. 2025. Cross-Platform Digital Discourse Analysis of the Israel-Hamas Conflict: Sentiment, Topics, and Event Dynamics. arXiv:2601.02367 [cs.CY] DOI: 10.48550/arXiv.2601.02367

Dan Bai and Yan Gu. 2026. Harnessing Big Data, Hindered by Bias: Evaluating TikTok Research API for Fair and Optimal Social Sciences. Social Science Computer Review (2026). DOI: 10.1177/08944393251413277

Steven Bird, Ewan Klein, and Edward Loper. 2009. Natural Language Processing with Python. O’Reilly Media.

Brasil. 2025. Lei n.º 15.211, de 22 de dezembro de 2025 – Estatuto Digital da Criança e do Adolescente. Diário Oficial da União. [link] Acessado em junho de 2026.

Fatmaelzahraa Eltaher, Rahul Krishna Gajula, Luis Miralles-Pechuán, Patrick Crotty, Juan Martínez-Otero, Christina Thorpe, and Susan McKeever. 2025. Protecting young users on social media: Evaluating the effectiveness of content moderation and legal safeguards on video sharing platforms. arXiv:2505.11160 [cs.SI]

Carlos Entrena-Serrano, Martin Degeling, Salvatore Romano, and Raziye Buse Çetin. 2025. TikTok’s Research API: Problems Without Explanations. arXiv:2506.09746 [cs.SI] DOI: 10.48550/arXiv.2506.09746

Diva Zahra Fakhirah and Fat’Hah Noor Prawita. 2026. Cross-Platform Sentiment Analysis of Indonesia’s LPDP Scholarship Policies: A Comparative Study Using IndoRoBERTa. In 2026 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT). IEEE, Bali, Indonesia. DOI: 10.1109/IAICT71158.2026.11620760

Deen Freelon. 2018. Computational Research in the Post-API Age. Political Communication 35, 4 (2018), 665–668. DOI: 10.1080/10584609.2018.1477506

Luis A. González-Camacho et al. 2025. Public opinion and emotional discourse: a study of YouTube comments on Mexican news in 2024. International Journal of Data Science and Analytics (2025). DOI: 10.1007/s41060-025-00936-3

Maarten Grootendorst. 2020. KeyBERT: Minimal keyword extraction with BERT. DOI: 10.5281/zenodo.4461265

Maarten Grootendorst. 2022. BERTopic: Neural topic modeling with a class-based TF-IDF procedure. arXiv preprint arXiv:2203.05794 (2022). DOI: 10.48550/arXiv.2203.05794

Jieun Kim et al. 2025. Unpacking College Students’ Mental Health Discourse through YouTube Comments: Insights from Topic Modeling and Sentiment Analysis. In Proceedings of the Association for Information Science and Technology, Vol. 62. DOI: 10.1002/pra2.1321

Chao Liu, Fang Fang, Xin Lin, Tianyi Cai, Xinchi Tan, Jintao Liu, and Xin Lu. 2021. Improving sentiment analysis accuracy with emoji embedding. Journal of Safety Science and Resilience 2, 4 (2021), 246–252. DOI: 10.1016/j.jnlssr.2021.10.003

Victor Martins, Sophia Eduarda V. Serafim, Lívia Caroline R. Pereira, Amanda F. Alves, Carlos H. G. Ferreira, and Jussara M. Almeida. 2025. A Misoginia no YouTube Brasileiro: Um Estudo de Caso sobre o Conteúdo Produzido pela Comunidade Red Pill. In Proceedings of the Brazilian Symposium on Multimedia and the Web (WebMedia). SBC, Rio de Janeiro, Brazil.

Leland McInnes, John Healy, and Steve Astels. 2017. hdbscan: Hierarchical density based clustering. Journal of Open Source Software 2, 11 (2017), 205. DOI: 10.21105/joss.00205

Leland McInnes, John Healy, and James Melville. 2018. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction. arXiv preprint arXiv:1802.03426 (2018). DOI: 10.48550/arXiv.1802.03426

João Olivetti and Philipe Melo. 2024. Through the Eyes of Instagram: Analyzing Image Content utilizing Meta’s Automatic Alt-Text. In Proceedings of the 30th Brazilian Symposium on Multimedia and the Web (Juiz de Fora/MG). SBC, Porto Alegre, RS, Brasil, 275–282. DOI: 10.5753/webmedia.2024.241695

Valeria Pedrouzo, Lucila Krynski, et al. 2024. More Skin, More Likes! Measuring Child Exposure and User Engagement on TikTok. arXiv:2408.05622 [cs.SI]

Rute Pina. 2025. Adultização: ’Vídeo de Felca conseguiu unir de Érika Hilton a Nikolas Ferreira e pauta cobrança por responsabilização das redes’. BBC Brasil. [link] Online; Acesso em 20 de junho de 2026.

Giovana Piorino, Vitor Moreira, Luiz Henrique Quevedo Lima, Adriana Silvina Pagano, and Ana Paula Couto da Silva. 2024. Análise de sentimentos de conteúdo compartilhado em comunidades brasileiras do Reddit: Avaliação de um conjunto de dados rotulados por humanos. In Proceedings of the Brazilian Symposium on Multimedia and the Web (WebMedia). SBC, Juiz de Fora, Brazil. DOI: 10.5753/webmedia.2024.242020

Rosario Polito et al. 2022. Hypersexualization of minors on TikTok: An exploratory study. American Journal of Sexuality Education 17, 4 (2022).

Juan Manuel Pérez, Mariela Rajngewerc, Juan Carlos Giudici, Damián A. Furman, Franco Luque, Laura Alonso Alemany, and María Vanina Martínez. 2023. pysentimiento: A Python Toolkit for Opinion Mining and Social NLP tasks. arXiv:2106.09462 [cs.CL]

Nils Reimers and Iryna Gurevych. 2020. Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. [link]

Marcus Vinicius Guerra Ribeiro, Clara Andrade Pimentel, and Philipe de Freitas Melo. 2024. Quando as Avaliações Viram Bombas: Explorando a Dinâmica do Review Bombing nos Jogos no Metacritic. In Proceedings of the Brazilian Symposium on Multimedia and the Web (WebMedia). SBC, Juiz de Fora, Brazil.

Michael Röder, Andreas Both, and Alexander Hinneburg. 2015. Exploring the Space of Topic Coherence Measures. In Proceedings of the Eighth ACM International Conference on Web Search and Data Mining (WSDM). 399–408. DOI: 10.1145/2684822.2685324

Jônatas H. dos Santos, Julio C. S. Reis, Philipe de F. Melo, João F. H. Olivetti, Thales H. Silva, Matheus Guimaraes, Glaucio de Souza, Marco A. G. Rodrigues, Marcos A. Goncalves, Fabricio Benevenuto, and Cristiano X. Lima. 2025. Caracterização da (Des)Informação das Narrativas sobre Vacinação em Grupos Públicos no WhatsApp no Brasil. In Proceedings of the Brazilian Symposium on Multimedia and the Web (WebMedia). SBC, Rio de Janeiro, Brazil.

Rebeca Suárez-Álvarez, Antonio García-Jiménez, and María L. Urbina Montana. 2023. Sexualising characteristics of adolescent on TikTok: Comparative study Great Britain–Spain. Convergence 29, 5 (2023), 1262–1282. DOI: 10.1177/13548565231187728

Ni Annie Yuan and Ho-Chun Herbert Chang. 2026. Motivation, Attention, and Visual Platform Design: How Moral Contagions Spread on TikTok and Instagram in the 2024 United States Presidential Election. arXiv:2602.02479 [cs.SI]