Modeling engagement of programming students using unsupervised machine learning technique

• Main Authors: FWA, Hua Leong, MARSHALL, Lindsay
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2018
• Subjects: unsupervised; machine learning; engagement; intelligent tutoring; sensors; Programming Languages and Compilers
• Online Access: https://ink.library.smu.edu.sg/sis_research/6971; https://ink.library.smu.edu.sg/context/sis_research/article/7974/viewcontent/Model_Engagement_Programming_2018_pvoa.pdf
• Institution: Singapore Management University

Engagement is instrumental to students’ learning and academic achievements. In this study, we model the engagement states of students who are working on programming exercises in an intelligent tutoring system. Head pose, keystrokes and action logs of students automatically captured within the tutoring system are fed into a Hidden Markov Model for inferring the engagement states of students. With the modeling of students’ engagement on a moment by moment basis, intervention measures can be initiated automatically by the system when necessary to optimize the students’ learning. This study is also one of the few studies that bypass the need for human data labeling by using unsupervised machine learning techniques to model engagement states.