Detection and assistance to students who show frustration in learning of algorithms

This paper presents a research work on the detection of students who show signs of frustration in learning activities in the area of algorithms, to then assist them with proactive support actions. Our motivation for the development of this work comes from students' difficulty in learning the concepts and techniques for building algorithms, which constitutes one of the main factors for the high dropout rates of computing courses. With the intent of giving a contribution to the reduction of such evasion, this research highlights the importance of considering students' affective states, trying to motivate them to study and work out their difficulties, with the assistance of computer systems. For research validation purposes, a tool was built to: a) infer the student's affective state of frustration while solving exercises of algorithms; b) detect signs associated with frustration, to provide resources to support student learning. Case studies were conducted with students of algorithms at the Faculty of Technology Senac Pelotas, in 2011 and 2012. The rules generated by the data mining software used to identify students' affective state of frustration, as well as an analysis of students' performance are presented in this article.