Autor(es):
Calado, Jorge Miguel da Silva
Data: 2017
Identificador Persistente: http://hdl.handle.net/10362/21765
Origem: Repositório Institucional da UNL
Assunto(s): Student engagement; Automatic emotion detection; Affective computing; Eye tracking; Facial expression recognition; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática
Descrição
Some of the biggest problems tackling Higher Education Institutions are students’ drop-out and academic disengagement. Physical or psychological disabilities, social-economic or academic marginalization, and emotional and affective problems, are some of the factors that can lead to it. This problematic is worsened by the shortage of educational resources, that can bridge the communication gap between the faculty staff and the affective needs of these students. This dissertation focus in the development of a framework, capable of collecting analytic data, from an array of emotions, affects and behaviours, acquired either by human observations, like a teacher in a classroom or a psychologist, or by electronic sensors and automatic analysis software, such as eye tracking devices, emotion detection through facial expression recognition software, automatic gait and posture detection, and others. The framework establishes the guidance to compile the gathered data in an ontology, to enable the extraction of patterns outliers via machine learning, which assist the profiling of students in critical situations, like disengagement, attention deficit, drop-out, and other sociological issues. Consequently, it is possible to set real-time alerts when these profiles conditions are detected, so that appropriate experts could verify the situation and employ effective procedures. The goal is that, by providing insightful real-time cognitive data and facilitating the profiling of the students’ problems, a faster personalized response to help the student is enabled, allowing academic performance improvements.
