A Multi-agent Modeling Framework for Psychology Process of Online Bidders
There are two major types of recent researches in e-auction: 1) Field study of consumer behavior 2) modeling. As agent-based simulation became a new tool to explain social networks, more and more researchers applied MAS in e-market field. However, most former jobs focus on improving the intelligent algorithm, but not psychology-oriented design of agent. Therefore, the agent-based framework proposed in this paper can simulate the dynamic decision process of online bidders, which is much closer to the real world. As a continuous computational experiment study of psychology framework proposed by Ariely and Simonson(2003), the Multi-agent based experiment was a three-stage model: Auction Entry decision, Middle phase of Auction, Conclusion of Auction. During the whole dynamic decision process, bidders behavior is both a time-dependent behavior and interactive behavior. Thus, Bidder agents decided whether or not enter a particular auction based on information collected from the market space. After enter a particular auction, endowment effect and reference price contribute to bidding decisions. Finally, in the conclusion of an auction, agent decided how much they want to win based on their feeling about losing versus wining.
Multi-agent Modeling Consumer Behavior Online Auction Computational Social Science
LIU Ruiyu XU Haoyue
Economics and Management School, Wuhan University, P. R. China
国际会议
The Fifth InternationalSymposium on Management of Technology(ISMOT07)(第五届技术与创新管理国际研讨会)
杭州
英文
1735-1738
2007-06-01（万方平台首次上网日期，不代表论文的发表时间）