Authors: Darina Dicheva, Trayan lliev
Addresses: Computer Science Department, Winston-Salem State University, 601 Martin Luther King, Jr. Drive, Winston-Salem N.C. 27110, USA. ' Faculty of Mathematics and Informatics, University of Sofia, 5 James Bouchier Str., Sofia 1164, Bulgaria
Abstract: A reactive microworld (RM) is an open-ended, free discovery, and reactive learning environment. When working the learner describes his goal, i.e. supplies a model of the intended result in terms of key concepts. RM recognises whether the key concepts are used correctly and, if not, reacts. The problem in this scenario is related to the high diversity characterising all dimensions of RM: tasks, solutions, key concepts usage, and feedback. It is not possible to pre-store in the system all this potential variety. Two basic ways can be applied to solve the problem: rule-based and case-based reasoning. The present work is an attempt to solve some of the diversity problems in the context of MYCROFT-M – a reactive microworld for debugging simple turtle graphics programs, which helps the student by suggesting appropriate edits. Since the space of possible corrections is very large we decided to employ the case-based reasoning (CBR) approach. In this paper we discuss a CBR extension of MYCROFT-M. We describe the developed CBR Debugger focusing on the main CBR themes: case representation, indexing, and retrieval and case adaptation.
Keywords: reactive learning environments; microworlds; case-based reasoning; program debugging.
International Journal of Continuing Engineering Education and Life-Long Learning, 1999 Vol.9 No.3/4, pp.237 - 248
Published online: 11 Dec 2009 *Full-text access for editors Access for subscribers Purchase this article Comment on this article