Research on knowledge based system for multipurposeful intelligent computer aided instruction

构建一个智能化的计算机辅助教学系统，需要一种推理机制、反映学生学习目标领域的领域理论、有效解决问题所需的元级知识以及所谓的学生模型。许多传统的实验ICAI系统通过应用相应的错误规则来模拟学生的错误操作，但当学生真正使用错误规则而不是正确的规则时，他们难以做出决策。我们的系统采用了另一种方法，在这种方法中，系统将试图通过基于解释的概括(EB)来获得方法，从而教会学生如何有效地解决问题。使用这种方法，当我们给出几个如何解决元级规则的例子时，系统可以提取元级规则。针对学生模型，我们采用真值维护系统(TMS)作为表示学生知识简明模型的基本工具，并用Prolog语言实现。

To construct an intelligent computer aided instruction system, an inference mechanism, domain theory which reflects the objective domain to be learned by students, meta-level-knowledge necessaryfor efficiently solving the problems, and so called student models. Many traditional experimental ICAI system model student's erroneous operation by applying the corresponding buggy rules, but they suffer form making decision when stdents really uses buggy rules instead of correct ones. Our system adopt another approach in which the system will try to teach students how to solve problems efficiently, by acquiring the method through explanation-based generalization (EB). With this approach, the system can extract meta-level rules when we give it several example on how they are solved. Concering student models, we have adopted Truth Maintenance System (TMS) as a basic tool for representating brief model of student's knowledge, realized it in Prolog.