Sergio Freitas brings data intelligence and academic risk to UnB's 9th Assessment Forum

Brasilia, November 9, 2026

Prof. Sergio Antônio Andrade de Freitas is one of the invited speakers at the 9th Assessment Forum of the University of Brasília (IX Fórum de Avaliação da UnB), on 16 November. He will speak on the panel “From Assessment to Action” with the talk “Data Intelligence in Institutional Assessment: identifying risks and supporting decisions”.

The forum is organised by UnB’s Institutional Self-Assessment Committee (CPA/UnB) and its Directorate of Institutional Assessment and Statistics (DAI/DPO/UnB). This year’s theme is “From assessment to action: evidence, good practices and results to transform the University”, and the event starts from one question: how can assessment results be turned into decisions, practices and real improvements for the University? The programme covers the administrative and academic management of academic units, the fight against student dropout, good practices and performance indicators.

The panel

The “From Assessment to Action” panel runs from 2 pm to 4 pm, moderated by Prof. Paulo Lima Junior. Alongside Prof. Sergio, the speakers are:

• Dr. Guilherme Viana, Director of Institutional Assessment and Statistics (DAI/UnB), on A taxonomy of dropout in higher education;
• Prof. Juliana de Freitas Dias, of the Directorate of Innovation for Undergraduate Teaching (DIEG/UnB), on From admission to retention: monitoring access.

The research behind the talk

The organisers’ invitation cites the professor’s track record in artificial intelligence, learning analytics, machine learning and educational technology, as well as his experience in managing undergraduate education at UnB, where he served as Dean of Undergraduate Education. Part of that track record is on the CEDIS website:

• Dropout and admission routes. The paper Formas de ingresso em perspectiva comparada: por que o SISU aumenta a evasão? O caso da UnB (in Portuguese) was published in 2021 in Avaliação: Revista da Avaliação da Educação Superior. It compares retention among students who entered the University through different admission routes.
• Predicting dropout risk. Supervised by the professor, Edmilson Cosme da Silva’s master’s dissertation Previsão da evasão acadêmica no ensino superior: o caso dos cursos de graduação presencial da UnB (2023, in Portuguese) applied machine learning to data from UnB’s on-campus undergraduate programmes. The systematic review of the factors that affect the prediction of retention and dropout , presented at the 56th Hawaii International Conference on System Sciences (HICSS 2023), and the undergraduate thesis Identificação de fatores que afetam a evasão no ensino superior (2022, in Portuguese) complete this strand.
• Learning indicators. The book Indicadores de Aprendizagem (Learning Indicators, 2023), written with the A3M Programme, received an honorable mention in the first edition of UnB’s Annual Award for Innovation in Undergraduate Teaching ( read the story ).
• Data for academic management. The academic resource allocation model applied at UnB received the third best paper award at IEEE ITHET 2024 ( read the story ).
• Data with privacy. CEDIS Learning Analytics brings together pseudonymised data to produce indicators without exposing students, and the project Predictive framework for student engagement and performance combined learning analytics and machine learning to anticipate risks in gamified courses.

This work belongs to the research lines Artificial intelligence, language processing, and data science and Learning technologies, gamification, and human factors .

Event details

9th UnB Assessment Forum: “From assessment to action”

• Date: 16 November 2026, 8 am to 6 pm. The “From Assessment to Action” panel runs from 2 pm to 4 pm.
• Venue: Rectorate Auditorium (Auditório da Reitoria), Darcy Ribeiro Campus.
• Event page: IX Fórum de Avaliação da UnB, on the DPO/UnB website (in Portuguese)
• Free registration: event form (in Portuguese)