Research On Transaction Credit Problem Of Unstructured Crowd Transaction Network

The platform e-commerce in china has developed rapidly in recent years,and has made remarkable achievements.However,at the same time,it has also exposed its inherent platform defects,such as lack of interaction,platform information isolation,lack of personalization while searching and matching,and lack of accuracy and resistance to aggression while evaluating transaction credit.There is still room for further improvement in the operational efficiency and international image of e-commerce market in china.With the development of Internet and Internet of things,individuals,enterprises,government departments and goods in physical space are all connected together,showing the development trend of information,unstructured and network.With the popularization and application of big data technology and intelligent technology,these agents have become more and more intelligent.And many intelligent agents,together with the ideas of their respective consciousness spaces,can be uniformly mapped to the information space with the help of the network.And then form a new generation of e-commerce form,called as crowd transaction network.Transaction credit is the premise of safe and credible transaction,which is especially critical to the crowd transaction.However,at present,there are some problems in e-commerce transaction credit models,such as the transaction credit measurement is one-sided,the accuracy is not high,the anti-aggression is not enough,and the transaction credit evaluation mechanism and the transaction credit management mechanism are not combined closely and so on.Therefore,aiming at the above problems,this paper studies the transaction credit problem of crowd transaction network.The main research work is as follows:(1)The transaction credit model of crowd transaction network is constructed.On the basis of analyzing the influencing factors of transaction credit,a transaction credit model is established.The model comprehensively considers the transaction credit impact factors,including evaluation score,time when the transaction took place,amount,transaction result,number of transactions and so on,among which the transaction result includes the number of failed transactions,transaction failure amount,which improves the accuracy of transaction credit evaluation.In order to avoid malicious attacks such as collusion evaluation and denigration evaluation in the network,this paper divides the recommendation nodes into two categories,constructing the recommendation credibility for the two types of recommendation nodes respectively,which improves the credibility and accuracy of the recommendation.The simulation experiment of the transaction credit model is carried out,and the experimental results verify that the transaction credit model has good accuracy and anti-aggression.(2)The transaction credit management mechanism of the crowd transaction network.After analyzing the transaction flow and network topology of crowd transaction network,the methods of saving,updating and searching credit data are constructed.Firstly,the credit data is stored in buyers and circle of friends,which not only balances the network load,makes full use of the overall computing ability of the network,but also solves the problem of single point failure.Secondly,a credit data search algorithm based on the combination of breadth and depth is proposed,optimizes the forwarding strategy of query request,and comprehensively considers the satisfaction of path usage,the response of credit data search and average response time,which makes full use of the wisdom of each node in the network.The simulation results show that the search algorithm can obtain the relevant credit data quickly and more,and reduce the network load effectively.