Identificação de gargalos em e-learnings gamificados e indicação dos erros mais frequentes para viabilizar e priorizar melhorias

Resumo

Due to the growth of e-learning and the recent changes on student learning preferences and behavior, the scientific community has proposed new approaches to support the application of active learning strategies and gamification. This work presents a solution to identify ``bottlenecks'' in courses, with the purpose of reducing effort of professors and students on the learning process. The solution includes data mining and clustering using Self-Organizing Map (SOM) to verify the course tasks with the highest error rates and then to group its errors by similarity. Thus it was possible to prioritize interventions at the points that most hinder students' progress. The approach was successfully applied in an industrial automation e-learning.