Research On Trust Mechanism And Algorithm Of Trade Selection In C2C Transaction Situation

Along with the rapid development of Internet in the 21st century, Business activities based on the Internet, such as electronic commerce, have gained increasing popularity, of which the C2C (Consumer to Consumer) electronic transaction burst into our sight. C2C transactions online bring convenience to both buyers and sellers. However, though reputation rules played a big role, trust crisis which was rare in face-to-face trading inevitably emerged due to the information asymmetry in trading, causing some potential trades aborting. Hence building up trust relationship between both parties is particularly important in the C2C electronic transactions. How to design a reasonable trust mechanism according to the trade actions becomes an important issue, which is mainly concerned in this paper. The key to design a trust mechanism is the measurement index, according to which traders can decide whether to trade or not. Besides, the platform can take some measures by supervising the indexes of traders. Mechanism usability depends on the calculation method of index and the accuracy of the trust mechanism. Existing research achievement mainly focuses on analyzing the influencing factors and constructing the trust model. The former aims at the concrete ways for building trust while the latter aims at keeping the trust relationship among two parties throughout [1]. These studies focus on the trust itself, and lack objective understanding of the transaction background.In order to establish a more practical and effective trust mechanism, data mining method is employed to model transaction history of existing platform so as to explore the trust mechanism, and then the behavior strategy selection is analyzed on C2C using repeated game theory. Finally, a trust mechanism based on trade selection algorithm is established and we focus on the compare with previous mining model.Firstly, the paper analyzes the internal relations between reputation rules and trust mechanism according to the effect of reputation rules on the trading platform for C2C. Secondly, based on the sample of an existed platform, the mining model of several users'attributes which weigh more in transaction is established using decision tree algorithm in data mining. The users are classified into different groups so that the character attributes of the platform under the trust mechanism can be accurately predicted. Then according to behavior strategies of traders, this paper establishes repeated game model of trading, and analyzes the behavior of speculative traders that adopt hybrid strategy using rational expectations in economics. As a result, the calculation method of traders'trust measurement index is designed through rationally targeting various trust related factors. Finally, a trust mechanism of GAOTS algorithm (Game Algorithm of Trade Selection) is established. Moreover, we use complex adaptive systems modeling tools--swarm to visually simulate GAOTS algorithm. Experiment results show the feasibility of the algorithm, and verify the algorithm can reflect reputation effect under certain conditions. Comparing to balance results of the existed platform under trust mechanism, GAOTS algorithm is incentive.By employing the research method of data mining, this paper models the trade attributes of C2C electronic market and the model can reflect the trust mechanism of the platform, while the trade selection game model established in this paper can help buyers select traders. Besides, the trust mechanism based on the model can accurately identify those non-credit transactions; when the electronic trading market is in a middle or large scale, the mechanism will improve the trading strategies of speculative traders, the satisfaction of buyers and market volume, so as to achieve mutually benefits.