Applying a Case-Based Reasoning System Development Tool in the Design of BDI Agents

Abstract

Most deliberative agents are designed based on the BDI architecture. Although the BDI architecture is uncomplicated and its decision process is easy to understand and execute the BDI architecture is difficult to implement and has no functionality of learning. Extending the architecture of BDI with CBR to develop the deliberative agent could make the development of agent easier compared to implementing a pure BDI architecture with the functionality of learning and adaptation. In this paper, we develop a domain-independent system development tool, JCBRDT. By using JCBRDT, the system developers could use CBR to develop a BDI agent, create a CBR system easily, and save the time on designing and maintaining the CBR system.