Inferencing Student Model for An Intelligent Computer Aided Instruction System in Civil Engineering

This research has developed an Inferencing Student Model and based on it an Intelligent Computer Assisted Instruction (ICAI) System. The Inferencing Student Model is based on Genetic Graph and Machine Learning Theory. It consists of a Master Expert Model, and an Inferencing Mechanism, which infers different Student Models from the one Master Expert Model. Based on the principle of learning by doing’ in the domain of static beam analysis, the system design provides greater flexibility in modelling student behaviour than previous systems. The system is implemented with PDC Prolog to run on PCs. The system, for the use of undergraduate civil engineering students, offers individualized learning and provides help during the learning process. Help is dynamically associated with the student’s needs and is controllable by the student. The learning is monitored and the learning record is available afterwards for the teacher to assess the student’s progress. The pilot study shows that the Inferencing Student Model is flexible, and successfully modelled the variety of student learning strategies encountered. The Inferencing Student Model also offers some improvability on site by a teacher. This feature was used to improve the system in response to weaknesses revealed by the pilot study.