PROPOSAL OF A PROBLEM-SOLVING PROCESS MODEL FOR LEARNING INTELLECTUAL PROPERTY LAW USING FIRST-ORDER PREDICATE LOGIC AND DEVELOPMENT OF A MODEL-BASED LEARNING SUPPORT SYSTEM

Here, we propose a methodology that uses first-order predicate logic to describe a problem-solving process model for learning intellectual property law, with the goal of developing a learning support system that uses this model. Specifically, the constructed learning support system describes the state of learner comprehension using first-order predicate logic, and uses differences in the logical structure of correct responses to uncover the comprehension state. In previous research, we defined a problem solving process model for learning intellectual property law based on one for physics. That model defines the problem-solving process as one that produces structures that include solutions represented as solution structures and constraint structures. When defining relational expressions between constraint and solution structures in physics, mathematical expressions can describe numerical relationships in natural law. While there are no quantitative relations in intellectual property law, legal clauses can be described as logical expressions (propositional logic), and thus can be used to define a problem-solving process model. The model in our previous research represented constraint structures as logical expressions, but conceptual relations between the subject and predicate in legal statements make propositional logic alone insufficient for representation. In this research, therefore, we use first-order and above predicate logic to represent subject predicate relations, and develop a learning support system that uses the resulting model. For problems that consider combinations of legal statements such as legal clauses and patent requirements, the developed learning support system retains first-order predicate logic representations that can allow direct calculation of problem structures with restrictions. Learners using the support system can likewise assemble solutions using first-order predicate logic. This realizes an inference mechanism that can calculate whether and how solutions contradict legal statements and problems. Namely, differences between correct and incorrect answers can be calculated and used as feedback to learners. This report describes the results of trials of the developed system.