An augmented Prolog programming environment for tutoring applications

This paper focuses on the design of an Augmented Prolog Programming Environment (APPE) for learning Prolog. It stresses the importance of being able to detect and classify errors present in the student program. In general, language compilers and interpreters can only detect syntax errors and can hardly diagnose semantic errors. It is conceived that a Prolog system can be enhanced considerably with the inclusion of an intelligent diagnoser and facilities for doing problem acquisition from the teacher and administering problems to the student. Unlike traditional CAI programs which support mainly teacher-centered tutorials, APPE facilitates the discovery mode of learning since it can cope with individual student programs. The design enables the diagnoser to classify Prolog program bugs into five categories: incorrect solutions, uncovered solutions, redundant solutions, nontermination, and invalid parameters. Further, the proposed diagnoser can identify various causes of nontermination bugs. Besides identifying the errors in a program, it can also provide remedial advice to the student programmer.