An implementation of a dynamic negotiation model for competitive and cooperative agents.

摘要

Agent technology evolved from number of disciplines such as object technology, distributed computing and artificial intelligence. Agents are supposed to autonomously engage in several types of dialogues to attain their design objectives. With the emergence of highly dynamic and uncertain e-commerce, there is currently a growing interest to develop negotiating software agents that can engage in buying and selling in a virtual marketplace. In this thesis report we propose a framework for automated dynamic negotiation among competitive and cooperative software agents. These agents are facilitated with BDI model of agency capabilities to accomplish dynamic negotiation. They make use of Case Based Reasoning techniques to learn from their previous experiences over a period of time. This framework also uses Ontology to equip agents with domain specific knowledge for reasoning purposes during dynamic negotiation. Our proposed framework suggests that it is possible to build software agents, which model limited aspects of dynamic negotiation.