An Automated Integrative Negotiation Agent

• Main Authors: Chu, Hung-Yang, 朱宏揚
• Other Authors: Chou, Seng-cho T.
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/28058513045163637812

碩士 === 國立臺灣大學 === 資訊管理研究所 === 88 === Negotiation is the most important stage in a commercial transaction. In today’s electronic market, the major automated negotiation mechanism is auction. Online auction simplifies the complicated negotiation behavior. Most online auction systems seen today negotiate solely over the product’s price rather than full range of values. These systems are so simple that users can easily understand how they work and delegate negotiation agents without much anxiety. However, a transaction can encompasses a broad range of attributes including price, quantity, quality and delivery schedules. Furthermore, the negotiation strategy could take into account the opponent’s behavior or environmental factors. Today’s online auction is considered to be too static that can’t truly reflect human’s negotiation behaviors. In this study, we establish an automated negotiation mechanism based mainly on the genetic algorithm technology. The system can negotiate over a set of given attributes. Furthermore, the agent searches a win-win proposal by giving up less important attributes in exchange for more important ones. Our experiments show that satisfactory results can be obtained.