Font Size: a A A

Study And Design On Multi-attributes Auction Modelin E-commerce Based On Intelligent Agents

Posted on:2008-05-20Degree:MasterType:Thesis
Country:ChinaCandidate:Q LiuFull Text:PDF
GTID:2178360215985057Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of software agent technology in E-commerce, agent-based electronic commerce is an inevitable trend. One of key technologies is the automated negotiation technology based on agent. Because of its publicity and competition, E-auction is becoming an important part in the research on electronic commerce. But there are two aspects weak in the current electronic auction. One is that the research and implementation of e-auction is focus on auction only for price. But in the B2B electronic commerce, participants do not only concern price but also more care cooperation relationship. The other is that the current electronic auction's automated degree is low. The auction participants have to be concerned with the auction course real time.According to the weakness of auction research and the need of electronic commerce, this article puts forward a multi-attributes electronic auction based on agent. The major research work in this thesis can be viewed as follow:Firstly, Construct a multi-attributes electronic auction framework based on mobile agent. On the basis of studying the prospect of application of mobile agent in electric commerce, this paper puts forward a framework base on mobile agent and multi-attributes electronic auction, discusses the process of auction and analyses some critical technical problems in front of us.Secondly, Put forward multi-attributes auction negotiation model in e-commerce based on mobile intelligent agents. The negotiation model puts forward a new multi-attributes auction protocol based on collaboration. This protocol strengthens the flexibility of system and adds the possibility of making a deal by virtue of permiting system to communicate some information for collaboration. The negotiation model uses BP neural networks to learn the buyer's prefer more accurately, then applies the genetic algorithm to design agent's bidding strategy and uses multi-attributes utility function approach to the winner determination problem under the condition of single supplier.At last, we research the implementation of multi-attributes electronic auction system. Then we design a multi-attributes auction system based on Agent, and introduce some relevant knowledge about Aglet which is the classic mobile Agent platform. The paper touches the frontier of international research and has wide application foreground.
Keywords/Search Tags:electronic commerce, multi-attributes auction, agent, genetic algorithm, neural networks
PDF Full Text Request
Related items