Development of a Learning Analytics Dashboard from Educational Data Mining Models

Abstract

This work proposes a development of Learning Analytics Dashboard for data analysis from the results of educational data mining, through the prediction of student performance. Thus, the general objective of this work is to create a visualization mechanism that aims to allow teachers to identify students whose predicted performance is unsatisfactory, providing information regarding the trajectory of students in the course and hence enabling the teacher to a pedagogical and motivational intervention. During the architecture development process, four stages were considered: awareness, where the focus is on visualizing data in different ways; reflection, whose objective is to assess the relevance of the data present in the visualizations, through questions asked to the teachers; sensemaking, which concerns the analysis and reflection of the answers given to the questions asked, creating new insights; impact, the final goal, which can produce changes in the course’s conduction by teachers, if they deem it necessary. Thereby, a tool was developed whose objective is to present student performance predictions in a friendly way for teachers, in order to contribute to the teaching-learning process and empower the interaction between students, teachers and resources in learning management systems.