Research Projects
AI Agency and Infrastructures in Social Research: Sociology, Anthropology, History, and Archival Science in the Age of Agents
(2025 – Present)
This project investigates the incorporation of AI agents into social science and humanities research, taking as its analytical axis the redistribution of agency among researchers, computational models, document bases, and digital infrastructures. The diagnosis is that the digital has ceased to be merely a set of sources, instruments, or auxiliary environments, and has become part of the very conditions of production, mediation, and validation of social, historical, anthropological, and archival knowledge. In this context, the expansion of large language models, retrieval-augmented generation systems, and autonomous agents demands a renewed epistemological critique capable of examining how such systems reorganize practices of interpretation, classification, search, synthesis, and documentation.
The central problem consists in understanding how the partial delegation of interpretive, classificatory, and documentary tasks to AI agents transforms the regimes of evidence, validation, and responsibility in research. The project’s hypothesis is that such systems do not replace the researcher but displace their methodological position: from direct and exclusive doing to the orientation, supervision, validation, and critique of distributed sociotechnical processes. Agency in research is thus understood as the result of a composition between human decisions, algorithmic operations, documentary curations, informational architectures, and audit protocols.
The investigation is organized into three articulated axes. The first examines the technical and epistemological foundations of large language models, embeddings, RAG systems, topic modeling, named entity recognition, and knowledge graphs, considering that every research technique incorporates theoretical assumptions, classificatory decisions, and methodological choices. The second analyzes AI agents as devices for action, planning, review, and monitoring in complex documentary environments, with emphasis on digital archives, multi-platform ecosystems, heterogeneous collections, and large masses of data. The third develops a critique of data, tools, and digital archives, problematizing the illusion of raw data, the opacity of technical mediations, the effects of information capture and organization infrastructures, as well as the limits of the mathematization of meaning in computationally driven research.
As a contribution, the project proposes a theoretical-methodological framework for the responsible, auditable, and epistemologically controlled use of AI agents in digital sociology, digital anthropology, digital history, and digital archival science. The ultimate objective is to offer conceptual and methodological instruments so that the participation of AI agents in knowledge production is not naturalized as neutral automation, but examined as a profound reconfiguration of the contemporary conditions of investigation in the social sciences and humanities.
Status: Ongoing | Students: Undergraduate (10) / Master’s (8) / PhD (6)
Team: Leonardo Fernandes Nascimento (PI) · Rosana Silva Moore Wedderburn · Bruno José Rodrigues Durães · Juciane Pereira de Jesus · Eric Brasil · Leticia Maria Costa da Nóbrega Cesarino · Daniel de Sena Bastos · Lidia Belas · Larissa Gualberto · Ricardo Sodré · Joyce Louback · Marcelo Freire Pereira de Souza
Multi-platform ecosystems and attacks on health information integrity: impacts, dissemination patterns, and mitigation strategies
(2024 – Present)
This project investigates how false, distorted, or inaccurate health information circulates, gains legitimacy, and negatively impacts public policies — especially childhood vaccination campaigns — within multi-platform digital ecosystems. The research adopts an interdisciplinary and methodologically mixed approach, combining complex network analysis, text mining, machine learning techniques, agent-based modeling, and multimodal linguistic analysis (text, image, video, and audio). It seeks to identify disinformation dissemination patterns, build monitoring tools such as interactive dashboards, and develop innovative health communication strategies. The initiative also includes actions aimed at scientific education, production of digital educational materials, and strengthening the capacity of health professionals, journalists, managers, and social movements to combat the infodemic. The project contributes to Brazil’s Unified Health System (SUS) by supporting evidence-based decisions, strengthening the bond with the population, and subsidizing epidemiological surveillance and public health communication actions.
Status: Ongoing | Students: Undergraduate (4) / Master’s (3) / PhD (4)
Team: Leonardo Fernandes Nascimento (Co-PI) · Maria da Glória Lima Cruz Teixeira (PI) · Jéfte Batista de Oliveira · Rosana Silva Moore Wedderburn · Bruno José Rodrigues Durães · Juciane Pereira de Jesus · Eric Brasil · Tarssio Barreto · Leticia Maria Costa da Nóbrega Cesarino · Daniel de Sena Bastos · Daniel Romero · Lidia Belas · Larissa Gualberto · Enzo Coelho Dias · Leonardo Thibau
Funder: CNPq
Digital Democracy: Analysis of disinformation ecosystems on Telegram during the 2022 Brazilian electoral process
(2022 – Present)
This project aims to establish a mapping framework for multi-platform ecosystems of Brazilian far-right groups on Telegram during the 2022 electoral process. The research adopts a mixed-methods perspective, articulating computational data analysis with socio-anthropological approaches. In a previous project, we analyzed the differences in dynamics between groups and channels, the prominence of users we call talkatives, and the significant presence of links to other mainstream social media platforms. In this second phase, the goal is to expand the collected data to include audio, video, and images shared in order to understand the narratives, values, grammars, and logics of action underlying users, groups, and channels.
Status: Ongoing | Students: Undergraduate (8) / Master’s (3) / PhD (6)
Team: Leonardo Fernandes Nascimento (PI) · Rosana Silva Moore Wedderburn · Juciane Pereira de Jesus · Eric Brasil · Tarssio Barreto · Jefte Batista · Paulo F. C. Fonseca · Leticia Maria Costa da Nóbrega Cesarino · Jorge Barbosa · Pedro Moraes · Daniel de Sena Bastos · Alan Delazeri Mocellim · Daniel Romero
Funders: InternetLab · CNPq
Telegram Observatory (collaboration with AI For Society, University of Alberta)
(2021 – 2024)
The Telegram social media environment is used extensively by extreme groups to communicate. We propose to develop the infrastructure at the U of Alberta that can follow specific Telegram channels in order to study communities, especially Alt-Right communities in the province. Further we propose to add search and analysis tools with AI capabilities to study the social media archives added.
Status: Completed | Students: Undergraduate (5) / Master’s (3) / PhD (3)
Team: Leonardo Fernandes Nascimento (PI) · Geoffrey Rockwell
Funder: AI for Society (University of Alberta)
Disinformation ecosystem and computational propaganda on Telegram
(2021 – 2021)
This project proposes to establish a mapping and multi-method analysis framework for far-right networks on Telegram, where a recent increase in user flow and the creation of groups and channels has been observed. Aiming to monitor possible computational propaganda activities and coordinated influence operations, the project combines computational analyses based on corpus linguistics and NLP with a mixed qualitative approach of discourse analysis and online ethnography.
Status: Completed | Students: Undergraduate (12) / Master’s (3) / PhD (4)
Team: Leonardo Fernandes Nascimento (PI) · Jéfte Batista de Oliveira · Rosana Silva Moore Wedderburn · Juciane Pereira de Jesus · Eric Brasil · Tarssio Barreto · Paulo F. C. Fonseca · Leticia Maria Costa da Nóbrega Cesarino · Vítor Mussa Tavares Gomes
Funder: CNPq
Analyzing the COVID-19 pandemic: social media, scientific controversies, and disinformation
(2020 – 2024)
The research aims to gather, process, analyze, and understand quantitative and qualitative data from digital social networks and newspapers about the COVID-19 pandemic. The analyses discuss themes in health anthropology, digital sociology, social studies of science, and ignorance studies (agnotology), using R, ATLAS.ti, Gephi, Voyant Tools, Iramuteq, among other tools.
Status: Completed | Students: Undergraduate (2) / Master’s (2) / PhD (8)
Team: Leonardo Fernandes Nascimento (PI) · Rafaela Zorzanelli · Francisco Ortega · Mylena Sousa Alecrim · Rosana Silva Moore Wedderburn · Juciane Pereira de Jesus · Eric Brasil · Tarssio Barreto · Jefte Batista · Paulo F. C. Fonseca · Leticia Maria Costa da Nóbrega Cesarino · Rafaela Rigoni
Digital Sociology: articulating CAQDAS, web scraping, and big data
(2019 – 2024)
Through modern data collection and analysis techniques (Atlas.ti, Nvivo, MAXQDA) combined with the R programming language, this research explores and improves the use of such technologies for social science research, with databases on historical images, public controversies, and the academic output of the social sciences on the Lattes platform.
Status: Completed | Students: Undergraduate (2) / Master’s (3) / PhD (2)
Team: Leonardo Fernandes Nascimento (PI) · Mylena Sousa Alecrim · Jéfte Batista de Oliveira · Patrícia da Silva Cerqueira · Juciane Pereira de Jesus · Marco Antonio Paranhos · Jorge Barbosa · Marcella Alencar
Sociology of diagnoses: the case of ADHD
(2017 – 2019)
The research aims to describe and analyze the process of expansion and legitimation of the Attention Deficit Hyperactivity Disorder (ADHD) diagnosis in Brazil through its diffusion in print and online media, using CAQDAS tools and databases of news articles and YouTube videos.
Status: Completed | Students: Undergraduate (3) / Master’s (1)
Team: Leonardo Fernandes Nascimento (PI) · Ive Cristine Pereira de Carvalho · Jéfte Batista de Oliveira · Juciane Pereira de Jesus · Gabriel Andrade
Digital labor in the 21st century: the case of Uber in Brazil
(2016 – 2018)
This project investigates the controversy surrounding Uber’s arrival in Brazil, articulating sociological, economic, and legal arguments as well as media coverage of the topic.
Status: Completed | Students: Undergraduate (6) / Master’s (2) / PhD (2)
Team: Leonardo Fernandes Nascimento (PI) · Ive Cristine Pereira de Carvalho · Jéfte Batista de Oliveira
Africa in the News: 140 years of the social construction of the African continent in the Brazilian press
(2016 – 2017)
Through critical discourse analysis of news about Africa in the newspapers Folha de S. Paulo and O Estado de S. Paulo, the research demonstrates how different news stories combined to produce racist and colonialist representations of Africa and Africans, using web scraping and ATLAS.ti.
Status: Completed | Students: Undergraduate (4)
Team: Leonardo Fernandes Nascimento (PI) · Ive Cristine Pereira de Carvalho
The multifaceted views of the press on the 30-year trajectory of the MST (1984–2014)
(2015 – 2019)
Analysis of the coverage by major Brazilian media (Folha de S. Paulo, O Globo, Estado de S. Paulo) about the Landless Workers’ Movement (MST) from its founding in 1984 to 2014.
Status: Completed | Students: Undergraduate (3)
Team: Leonardo Fernandes Nascimento (PI)
Funders: CNPq · FAPESB
Cultural studies in Brazil: a possible map
(2015 – 2019)
Mapping the profile of cultural studies in Brazil based on papers presented at the editions of the Meeting of Multidisciplinary Studies in Culture (ENECULT).
Status: Completed | Students: PhD (4)
Team: Mariella Pitombo Vieira (PI) · Leonardo Fernandes Nascimento · Lindinalva Rubim · Delmira de Souza
Dengue: 140 years of the social construction of an epidemic
(2015 – 2017)
Analysis of 140 years of news coverage related to dengue in the three main Brazilian newspapers, describing the intertwining of medical and public policy discourses.
Status: Completed | Students: Undergraduate (2)
Team: Leonardo Fernandes Nascimento (PI) · Lara Rosa Meirelles Barros · Ive Cristine Pereira de Carvalho
The Bolsa Família Program in the Brazilian press: ten years (2003–2013) of a public controversy
(2015 – 2017)
Mapping ten years of public debate in the Brazilian press about the Bolsa Família Program, emphasizing valence, framing, arguments, and representations in the newspapers Folha de S. Paulo, O Globo, and Estado de S. Paulo.
Status: Completed | Students: Undergraduate (2) / Master’s (1) / PhD (1)
Team: Leonardo Fernandes Nascimento (PI) · Ive Cristine Pereira de Carvalho · Jéfte Batista de Oliveira · Juciane Pereira de Jesus
Funders: CNPq · FAPESB
“Becoming a modern woman”: feminine sanitary pads and civilizing processes
(2015 – 2017)
One hundred years of news articles describing the arrival of feminine sanitary pads in Brazil and the sanitary and social justifications for their use, analyzing the construction of an “affective economy” (Triebhaushalt) regarding how women should deal with menstruation.
Status: Completed | Students: Undergraduate (2)
Team: Leonardo Fernandes Nascimento (PI) · Lara Rosa Meirelles Barros · Ive Cristine Pereira de Carvalho
ADHD: the controversy around a diagnosis
(2014 – 2019)
Description and analysis of the process of expansion and legitimation of the ADHD diagnosis in Brazil, based on databases of news articles (audio, video, text, and images) analyzed with CAQDAS tools.
Status: Completed | Students: Undergraduate (1)
Team: Leonardo Fernandes Nascimento (PI) · Lara Rosa Meirelles Barros
Development of a Sustainable Monitoring and Evaluation System for National Guidelines for Comprehensive Health Care
(2013)
Production of evidence to support the work of the Adolescent and Youth Health Care Area (ASAJ/DAPES/SAS/MS), influencing the implementation of the National Guidelines for Comprehensive Adolescent and Youth Health Care.
Status: Completed | Students: Undergraduate (2) / Master’s (1) / PhD (1)
Team: Marcelo Rasga Moreira (PI) · Leonardo Fernandes Nascimento · João Feres Júnior
Research on the construction of identity in representations of autism in the Brazilian press
(2012 – 2017)
Mapping the public understanding of autism in Brazil based on narrative analysis of media representations, proposing the narrative paradigm for thinking about the construction of autistic identity and its political implications.
Status: Completed | Students: Undergraduate (3) / PhD (1)
Team: Francisco Ortega (PI) · Leonardo Fernandes Nascimento · Clarice Rios · Rafaela Zorzanelli
Monteiro Lobato and political correctness
(2011 – 2013)
Analysis of the journalistic corpus (print and television) surrounding the complaint to SEPPIR about racist content in Monteiro Lobato’s Caçadas de Pedrinho, aiming to understand the processes of construction of public controversies in Brazil and the role of the media in shaping them.
Status: Completed | Students: Undergraduate (2) / PhD (1)
Team: João Feres Júnior (PI) · Leonardo Fernandes Nascimento · Zena Winona Eisenberg