Implements a diagnostic intelligent agent for problem solving in instructional systems

Abstract

Problem solving is one way to present thinking style of human beings. Most problems should be solved in some particular sequences. However, no matter how smart learners may possibly make mistakes while solving a problem. In order to decrease the mistakes making and increase the learning efficiency, a diagnostic intelligent agent is implemented in this paper to achieve the goal. Such a diagnostic intelligent agent is designed based on diagnosis PSN (DPSN), which modified from PSN. This paper takes one-variable linear equation (OVLE) as an example to accomplish the experiment instructional system in which a diagnostic intelligent agent, which is so-called OVLER, is involved. By analyzing the problem solving operations and representing the problem states with graph structure, a diagnosis PSN (DPSN) is designed (Because of the page limitations, the detailed lemmas and its proofs are ignored here.) According to the DPSN, a diagnostic intelligent agent-OVLER is finally implemented in the IVC (Internet Virtual Classroom) for demonstration purpose. From the problem solving states solved by learner, OVLER will detect the factors that cause errors and lead the learners to find the true answer.