LearnOOP: An active agent-based educational system

Abstract

Computer-based learning systems have become increasingly important. However, such self-study systems still lack the aspect of a real-life classroom and suffer due to their passive nature. They are often dull and plodding. This paper describes the LearnOOP, an active distance learning system for learning object-oriented programming. LearnOOP has several features: (1) active, i.e. the system is like a human teacher, who monitors the student's activities and gives guidance when necessary, (2) interactive, i.e. the student can ask the system questions and the system returns intelligent answers and (3) collaborative and co-operative, i.e. there are facilities for students and teachers to collaborate and co-operate with each other. In this paper, a powerful conceptual framework has been presented using a knowledge representation language, Telos, to facilitate the formal thinking about the tools and techniques used in designing, developing and using co-operative computer-based learning systems. The architecture of LearnOOP consists of three layers, the educational agent layer, the knowledge server layer and the repository layer. By deploying repository technology, the architecture provides a means for multiple intelligent educational agents to conduct active interactions between students and the system, between students and teachers and among students. The LearnOOP prototype demonstrates the need for multiple intelligent agent co-operation and collaboration in future active and co-operative learning systems and illustrates the feasibility for practical future design and deployment of such systems.