Research Of Matchmaking Trading Model Based On The Electronic Market

The application of electronic commerce must be supported by trading model and trading platform.With the development of network technique,more and more professional trading platforms have appeared,but the investigation of trading model falls behind relatively.The development of electronic commerce require to refine the traditional trading model,construct an integrated electronic commerce environment,provide an intelligent,secure and efficient scheme for enterprise to trade in Internet urgently. At present,the investigation of trading model in the native still rest on realising the traditional trading model using information skills,little work has been done in lucubrating the business model and exercising successfully.Although foreign investigation has gained some successes and brought forward some typical trading models such as fixed price selling ,biliteral multi-step negotiation and all kinds of auction,there still exists some problem such as information overloading or resource mazing in these models,which aggravate users' burden.How to realise an efficient,intelligent trading model based on the multi-agent technique has become an important task for the research of electronic commerce.In this paper,we study the trading model based on the project "the network market",which was implemented by the ChongQing electronic commerce Inc and us.Aiming at the shortage of trading model in "the network market",we employ game theory and multi-criteria decision theory, introduce the matchmaking schema based on price and quantity into electronic commerce application,bring forward a market matchmaking trading model including five phases:market matchmaking ,reinforce learning ,biliteral negotiation, contract signing,contract executing.The main work and conclusion as follows:Considering the behaviour of multi-buyers with multi-sellers, we realise a matchmaking model based on the price and quantity through double auction mechanism under discriminatory and non- discriminatory price situation,analyse the incentive compatibility and competitive equilibrium of the mechanism.In order to enhance the market efficiency,the paper advance a learning mechanism for the double auction,which was based on the Roth_Erev's three parameters learning model and modified slightly .By virtue of the trading history,the agent can learn by itself.The negotiation model is a necessary complementary for the matchmaking . Whenboth sides need to decide on other pending items,they can launch a biliteal multi-rounds negotiation.To ensure the efficiency and justice of negotiation,we adopt a improving directions negotiation model and a gentic algorithm negotiation model ,analyse and prove the model's feasibility and correctness.Then,we introduce the contract signing and contract executing process briefly.The conclusion and prospect are in the end .