Evaluating a CBR System for Predicting Student Performance

S.T. Karamouzis (USA)

Keywords

Casebase Reasoning, Prognostication, Expert System, Application in Education

Abstract

Research in predicting student performance in an educational setting resulted in the development of an intelligent system that uses case-based reasoning in order to forecast student class performance. The system draws conclusions on the basis of similarities between a student's current class performance and the performance of other students that attended the same class. This paper evaluates the intelligent system and presents the results achieved. A Turing-like comparison, where the system's performance is compared and contrasted with the prediction abilities of human instructors, places the achieved results in perspective. Findings of the evaluation indicate that in comparison to humans the system outperformed non-expert instructors. Educators may develop similar systems that are customized to the structure of their own classes and are capable of assisting them in advising students on their class progress way before it is too late for the student.

Important Links:



Go Back