Bruno Penteado

Followers

122

Following

Mentions

Public Views

A computer scientists trying to figure out how technology can be used to improve our lives.

less

Yuksel Goktas

Ataturk University

Thom Brooks

Durham University

Richard Bellamy

University College London

Martin Oliver

UCL Institute of Education

Ramesh Chander Sharma

Ambedkar University Delhi

Armando Marques-Guedes

UNL - New University of Lisbon

Patrick Lowenthal

Boise State University

George Kakarontzas

UNIVERSITY OF THESSALY, GREECE

Andrew W Wilkins

Goldsmiths, University of London

Bruno Silva

University Salvador - UNIFACS

Interests

Uploads

Papers by Bruno Penteado

Autenticação biométrica de usuários em sistemas de E-learning baseada em reconhecimento de faces a partir de vídeo

Aleph, Jul 27, 2009

Saúde digital e a plataformização do Estado brasileiro

Ciencia & Saude Coletiva, Jul 1, 2023

Este é um artigo publicado em acesso aberto sob uma licença Creative Commons

Download

Guidelines for the Application of Data Mining to the Problem of School Dropout

Springer eBooks, 2022

Mapping the dynamics of medical informatics: a bibliometric analysis of the scientific field

Reciis, Nov 1, 2021

Educational data mining for understanding students' behavior in a virtual learning environments: a case study

Meeting abstracts, 2017

As redes de conhecimento dos documentários online: uma análise do documentário The Power of Gold na plataforma Twitter

Dimensões subjetivas na Saúde Digital

Liinc em Revista, Nov 30, 2022

Marcelo Fornazin c, e RESUMO: As heterogêneas e contraditórias estruturas do capitalismo global v... more Marcelo Fornazin c, e RESUMO: As heterogêneas e contraditórias estruturas do capitalismo global vêm moldando de diferentes formas os significados sociais acerca da Saúde Digital e, com isso, também seu desenvolvimento e regulação. Para uma melhor compreensão destas questões, nesta revisão de literatura destacamos que tais estruturas contêm certas dimensões subjetivas que precisam ser investigadas de forma mais detida. Para isso, analisamos como a Saúde Digital se torna um espaço tanto de expansão das potencialidades humanas quanto para o acirramento dessas contradições estruturais. Discutimos como os modos de governamentalidade capitalista seriam atualizados junto a esses movimentos, moldando subjetivamente patologias, pacientes e o cuidado em saúde. Analisamos como esses processos podem se desdobrar em técnicas comportamentais que podem ser embarcadas em dispositivos e aplicativos de saúde e bem-estar, bem como suas consequências e riscos potenciais, o que exige atenção regulatória com participação social.

Download

Design propositions for the critical analysis of educational machine learning-based applications using emancipatory pedagogy

Anais do XXXIII Simpósio Brasileiro de Informática na Educação (SBIE 2022), Nov 16, 2022

In education, machine learning applications provide support and analytics insights to students, t... more In education, machine learning applications provide support and analytics insights to students, teachers, and administrators. However, not all of us are treated equally by these technologies. The algorithmic bias may reflect unequal opportunities for individuals based only on their demographic data. Following a design science research approach, we investigate multiple sources of bias in the machine learning pipeline and use emancipatory pedagogy as kernel theory to elaborate design propositions to mitigate this problem. We correlate the sources of bias with potential actions, providing theoretical lenses to handle bias throughout the development of intelligent educational systems. These principles should provide researchers with a critical analysis of the development of intelligent systems in education. Resumo. Na educação, aplicações baseadas em aprendizado de máquina fornecem suporte e insights analíticos a alunos, professores e administradores. No entanto, nem todos nós somos tratados igualmente por essas tecnologias. O viés algorítmico pode refletir em oportunidades desiguais para os indivíduos com base apenas em seus dados demográficos. Seguindo uma abordagem de pesquisa em design science, investigamos múltiplas fontes de viés no pipeline de aprendizado de máquina e usamos a pedagogia emancipatória como teoria kernel para elaborar propostas de design para mitigar esse problema. Correlacionamos as fontes de viés com ações em potencial, fornecendo lentes teóricas para lidar com o viés no desenvolvimento de sistemas educacionais inteligentes. Esses princípios devem fornecer aos pesquisadores uma análise crítica do desenvolvimento de sistemas inteligentes na educação.

Download

Detecting Behavioral Trajectories in Continued Education Online Courses

2019 IEEE 19th International Conference on Advanced Learning Technologies (ICALT)

Continuing education online has been steadily increasing worldwide and its students' profiles... more Continuing education online has been steadily increasing worldwide and its students' profiles tend to be different from those in other formal modalities. Hence, it is relevant to understand the behavior of these students along their course trajectories. In this work, we investigate a specialization online course involving hearing healthcare professionals (n=96), to discover different navigation strategies and the factors which may influence it. To that end, we applied sequence analysis techniques to interpret different navigation patterns and selected demographic covariables which may influence these behavioral trajectories. Three clusters were found, with subtle differences among them and three variables with greater predictive power: age, professional experience and region.

Towards a Perspective to Analyze Emergent Systems in Health Domain

Lecture Notes in Computer Science, 2022

Toward Better Outcomes in Audiology Distance Education: An Educational Data Mining Approach

American Journal of Audiology, 2018

PurposeThis article introduces concepts and a general taxonomy used by the educational data minin... more PurposeThis article introduces concepts and a general taxonomy used by the educational data mining (EDM) community, as well as examples of their applications, with the aims of providing audiology educators with a referential basis for developing this area.MethodA narrative review was carried out to present an overview of EDM and its main methods. Some of these methods were exemplified with analysis of real data from an Internet-based specialization course on pediatric auditory rehabilitation.ResultsThe review introduced EDM main concepts and applications and described methods from its area. Real data examples illustrated EDM use to predict interpersonal help-seeking, model interpersonal interaction, analyze students' trajectories within a course's module, and understand how students approached group assignments. Some of the insights provided by EDM to support teaching and learning processes were also described.ConclusionsEDM methods offer new tools to discover knowledge from...

Download

The Regulation of Artificial Intelligence in Healthcare

Proceedings of the Conference on Computers and People Research

The new capabilities that emerged with artificial intelligence also target healthcare settings. N... more The new capabilities that emerged with artificial intelligence also target healthcare settings. Nevertheless, they carry potential benefits and challenges that need to be addressed appropriately. In this paper, we mapped the current status of scientific debate on two key regulation aspects of AI in healthcare: the objects of regulation and the affected actors. Data-related issues are the most researched, such as collection, storage, and privacy. In addition, nearly half of the papers targeted policymakers with challenges, comparative analysis, and governance models. Based on these results, we propose a scheme to organize these aspects and discuss implications for AI management, policymaking, and research in this field. CCS CONCEPTS •Social and professional topics → Computing / technology policy →Medical information policy → Medical technologies• Applied computing →Life and medical sciences → Health informatics

Download

An architecture for monitoring public educational policies based on big open linked data

Anais do XXXIII Simpósio Brasileiro de Informática na Educação (SBIE 2022)

Open government data (OGD) is increasingly being deployed in many countries. It is often consider... more Open government data (OGD) is increasingly being deployed in many countries. It is often considered a pillar of democracy by providing society with transparency and accountability for public policies and actions. However, the role of OGD in policy-making monitoring has been explored only marginally due to its complexity and inter-organizational boundaries. Big Open Linked Data (BOLD) is a recent technological approach where data from multiple institutions with different data governance policies must be integrated to achieve a common goal. In this approach, linked data is essential in providing tools and practices to merge and analyze these data. In this work, we present a BOLD infrastructure to support the monitoring of an educational public policy using a linked data approach. The design science research methodology (DSR-M) was followed in this study. We chose the PNE (National Education Plan) as an illustrative scenario since it sets out goals for national education at different l...

Download

A Method Proposal for Implementing Accessibility in Desktop Applications for Visually Impaired Users

Currently, little is said about the accessibility-oriented desktop applications. In the case of t... more Currently, little is said about the accessibility-oriented desktop applications. In the case of this study, problems related to the application structure, which has characteristics of a legacy software, brings challenges that hinder the implementation of accessibility. This article shows an implementation of accessibility by applying the concepts of web standards in desktop applications, addressing factors such as controlling the events of Flash components to make data accessible to screen readers and communication between the layers of user interface and business taking into account information accessibility through the use of technology MSAA.

Download

Evaluating Temporal Predictive Features for Virtual Patients Feedbacks

Proceedings of the 19th ACM International Conference on Intelligent Virtual Agents, 2019

One key challenge to create believable embodied conversational agents (ECA) is to produce engagin... more One key challenge to create believable embodied conversational agents (ECA) is to produce engaging behavior-and feedbacks (short verbal, vocal and gestural reactions produced when hearing the main speaker) play an important role. In this paper we propose a machine learning-based model for multimodal feedbacks. The goal is to learn, from a corpus of human-human interactions, when a virtual agent should display a feedback along with its type. And to be feasible, an important aspect is to be able to process them in real time, using reliable features. For this purpose, we used random forests with different features, using annotated corpora of taskoriented interactions. Our case study is the context of training doctors to break bad news to a patient (played by an actor or by the ECA). The performance of the method highlights the capacity to predict verbal and non-verbal feedbacks based on a small number of features characterizing temporal information, in particular, the silence and the position of the last feedback. CCS CONCEPTS • Computing methodologies → Artificial intelligence; Intelligent agents; Feature selection.

Download

Modelo de infraestrutura para publicação de dados abertos educacionais conectados de qualidade

Anais do X Workshops do Congresso Brasileiro de Informática na Educação (WCBIE 2021), 2021

Os dados abertos disponibilizados por governos em todo o mundo têm sido usados para o desenvolvim... more Os dados abertos disponibilizados por governos em todo o mundo têm sido usados para o desenvolvimento de pesquisas, serviços e políticas públicas, gerando insights e embasando ações. Um desafio atual é a dificuldade em conectar diferentes fontes desses dados para realizar cruzamentos complexos, devido a problemas como interoperabilidade e alinhamento semântico. Esta tese propôs um arcabouço de artefatos que subsidiam a publicação de dados abertos governamentais educacionais, dotados de mecanismos de controle de qualidade e de acordo com as melhores práticas de compartilhamento de dados na Web. Foram desenvolvidos três artefatos tecnológicos, sob a epistemologia da Design Science Research. Juntos, triangulam para tratarem o problema de modo mais completo do ponto de vista tecnológico. Foram conduzidos i) um estudo de caso sobre o problema da rotatividade docente, unindo diferentes fontes de dados e visualizando as movimentações e ii) um quasi-experimento em que os sujeitos publicaram...

Download

Utilização de notas escolares para predição da nota ENEM em ciências humanas

RENOTE, 2021

O presente trabalho explora a hipótese de predição da nota ENEM em ciências humanas a partir de d... more O presente trabalho explora a hipótese de predição da nota ENEM em ciências humanas a partir de dados pedagógicos oriundos de notas de avaliações escolares de estudantes de ensino médio de um colégio particular em São Paulo. A partir da análise, foi possível a predição da nota ENEM de ciências humanas no decurso do 3º ano do ensino médio com uma correlação (r de Pearson) de 60,2% e RMSE de 39,7 pontos na escala ENEM. O modelo apontou, ainda, conteúdos relevantes fortemente correlacionados com o desempenho ENEM já desde o 1º ano do ensino médio.

Download

An analytics approach to investigate teacher turnover

Anais do XXXII Simpósio Brasileiro de Informática na Educação (SBIE 2021), 2021

Some educational problems embed spatial and temporal complexities, and the aggregation of these d... more Some educational problems embed spatial and temporal complexities, and the aggregation of these data may cause contextual information to be lost. One such example regards teacher turnover, which impacts directly the students' learning processes. In this work, we adopted an observational cross-sectional methodology, using visual analytics techniques to identify complex patterns in the mobility data of teachers in public schools from the city of São Paulo between 2016 and 2017. For this, we used education open data from the Brazilian government, which maps which teachers teach in which schools through a yearly school census. In addition, we sought to understand which are the main factors that, along with institutional rules, influence this sort of decision. To contextualize the main factors, we used synthetic indicators developed by the Brazilian government to identify different motivation clusters that may influence teachers' decisions to move to another school. As result, we...

Download

Detecção de inovações tecnológicas na evolução da informática educacional no Brasil

Anais do XXXII Simpósio Brasileiro de Informática na Educação (SBIE 2021), 2021

Inovações em um campo de conhecimento são geralmente cíclicas e dependentes da infraestrutura tec... more Inovações em um campo de conhecimento são geralmente cíclicas e dependentes da infraestrutura tecnológica desenvolvida. Neste trabalho buscamos caracterizar as contribuições da informática na educação ao longo dos anos, evidenciando ciclos de inovação nas pesquisas feitas na área no Brasil. Para isso, aplicamos técnicas bibliométricas na principal revista nacional deste assunto – a Revista Brasileira de Informática na Educação (RBIE) entre 1997 e 2021. Três períodos homogêneos de pesquisas foram detectados, além de nove agrupamentos de temas, demonstrando a evolução desses temas ao longo dos períodos.

Download

Prediction of Interpersonal Help-Seeking Behavior from Log Files in an In-Service Education Distance Course

Lecture Notes in Computer Science, 2018

We propose a machine learning approach to automate the estimation of the interpersonal help-seeki... more We propose a machine learning approach to automate the estimation of the interpersonal help-seeking level of students in an online course, based on their behavior in an LMS platform. We selected behavioral and performance features from the LMS logs, using forum and wiki variables in the context of a professional development course in audiology rehabilitation (N = 93). Then, we applied different state-of-the-art regression algorithms to predict their responses, using student-level cross-validation in the training set and evaluated the resulting models in a separate test set. As result, we had approximately an error of one point with our model, on average. We discuss some deviant cases and how this information can be used to inform tutors in online courses.