A reusable model of causal graph in qualitative reasoning

摘要

This paper proposes a new data reuse model for graph-based simulation in qualitative reasoning. Qualitative-based simulation is one of important and traditional tools to analyze a dynamic system and to initiate knowledge and principles for students in artificial intelligence. To simply make a model and run a simulation, the graph-based qualitative simulation is a promising method to enhance the efficiency of making graph model, correctness of the model, and understanding the simulated essence. When a simulation model is developed and given the condition/assumption of nodes and arcs in actual use of simulator, the model is completed by referring from our defined set of components. Further, we propose a method to make a graph model based on tree-structured chart and a reuse method of sub-graph in the simulation graph model. The contributions and advantages of this paper can be described as follows. Our proposed making graph method and reuse method can easily make the simulation model. The graph can be also checked whether there are error by oversight or not, since the model can be shown visually on the implemented simulator.