Kit-Build Concept Map for Automatic Diagnosis

摘要

In this paper, we describe a framework of Kit-Build Concept Map (we call it as KB map) that can diagnose. The task to make a concept map is divided into two sub-tasks: 1) "segmentation task" where parts of the concept map are extracted and 2) "structuring task" where the extracted parts (kit) are connected into a map. In the framework of KB map, an ideal concept map (goal map) is prepared by a teacher or an expert at first, and parts are generated by decomposing the goal map. The parts are provided to learners, and then the learners build concept maps (learner maps) by connecting the parts. Since the same parts are used both in the goal map and learner maps, it is possible to diagnose the maps by comparing them. This paper mainly explains a practical flow of KB map building.