The issue of school failure and dropout is persisting and alarming as it presents repercussions to society on the whole. Emotion detection, also known as affective computing, is an evolving field of research focused on enabling intelligent systems to detect, perceive, understand, and interpret human emotions. This has great potential in many fields, including education. This dissertation presents an emotion detection approach through tools that were not developed for that purpose. By annotating a Likert-Scale questionnaire, student emotional traits were extracted and analyzed, including correlation to students’ risk of failing. Additionally, four machine learning models were tested and Support Vector Machine showed the best performance in accurately predicting student emotions. This approach was applied to a sample of 845 students from northern Portugal, within the context of the the Northern Regional Operational Programme (NORTE 2020), under Portugal 2020 within the scope of the project “Hello: Plataforma inteligente para o combate ao insucesso escolar”, Ref. NORTE-01- 0247-FEDER-047004, which aimed to develop a digital platform comprising a conversational agent and an education intelligence system capable of predicting students’ risk of failure, allowing for a timely detection and consequent action. Findings report that anxiety was the emotion that presented the highest scores, with male students scoring slightly higher than female students for this particular emotion. It was also found that there was a gradual decrease in positive emotions (happiness, trust, optimism, interest) as students progressed from the fifth to the ninth grade, with the eighth grade showing the lowest scores overall, though with positive emotions declining more than negative emotions (boredom, anxiety, distraction, shame). Students at low risk of failure presented higher scores for all emotions compared to other risk clusters, and higher optimism was associated with lower failure risk. This emotion detection model presents itself as an approach to combating school failure, contributing to the identification of students emotional traits.