Abstract: Performance of learning can be enriched with proper and timely feedback. This paper proposes a solution based on a Bayesian network in machine learning that can examine and judge students' written response to identify evidences that students fully comprehend concepts being considered in a certain knowledge domain. In particular, it can estimate probabilities that a student has known concepts in computer science at different cognitive ability levels in a sense. Thus, the method can offer learners personalized feedbacks on their strengths and shortcomings, as well as advising them and instructors of supplementary education actions that may help students to resolve any lacks to improve their knowledge and exam score.
Authors: Phuoc Hung Pham and Javed I. Khan (Kent State University, USA)
Email: hung205a2@gmail.com, javed@kent.edu