In order to reflect the subjective preference and reduce intervention of human in E-Commerce, this work presents a model using fuzzy logic to support multi-agent negotiation. In the model, we used fuzzy logic to describe human's subjective preference, calculated evaluation function to the preference of single issue, and obtained the global preference using utility function with the differ weight of these issues. Meanwhile, we used urgent degree to represent human factors, and proposed a new dynamic concession strategy based on feedback learning. One agent should consider not only its interest but also other offer when proposing counter offer. After we analyze experimental results, the presented negotiation agent model makes negotiators achieve mutual benefits and decreases human interference effectively.
展开▼
