An Agent Negotiation Skill Algorithm Based on Interactive Micro-adjustment of Proposal in Electronic Transactions

摘要

Negotiation strategy is an important research branch of agent-based automatic negotiation. Under the background of bilateral multi-issues negotiation, a novel negotiation skill algorithm by interactive micro-adjustment and multi-rounds approximation is proposed, which classifies both parties' proposal issues into two divisions-interest point set and disinterest point set, and combined with the historical experience data conducts in-depth negotiation process according to the composite relationship of both interest level of consistency proposal issues and difference of issue values. At the same time concern degree of opponent about proposal issues is given and updated by a fuzzy theory-based prediction function, which can realize the purpose of learning negotiation opponent. Finally, the application and analysis of the proposed negotiation skill algorithm are demonstrated by a numerical experiment.