Feedback improvement in automatic program evaluation systems

Abstract

Automatic program evaluation is a way to assess source program files. These techniques are used in learning management environments, programming exams and contest systems. However, use of automated program evaluation encounters problems: some evaluations are not clear for the students and the system messages do not show reasons for lost points. The author proposes several ideas for possible improvements in black box testing, which can lead into better service for the users of automatic evaluation systems. © 2010 Institute of Mathematics and Informatics, Vilnius.