Abstract
PurposeCurrent models of transaction credit in the e-commerce network face many problems, such as the one-sided measurement, low accuracy and insufficient anti-aggression solutions. This paper aims to address these problems by studying the transaction credit problem in the crowd transaction network.Design/methodology/approachThis study divides the transaction credit into two parts, direct transaction credit and recommended transaction credit, and it proposes a model based on the crowd transaction network. The direct transaction credit comprehensively includes various factors influencing the transaction credit, including transaction evaluation, transaction time, transaction status, transaction amount and transaction times. The recommendation transaction credit introduces two types of recommendation nodes and constructs the recommendation credibility for each type. This paper also proposes a “buyer + circle of friends” method to store and update the transaction credit data.FindingsThe simulation results show that this model is superior with high accuracy and anti-aggression.Originality/valueThe direct transaction credit improves the accuracy of the transaction credit data. The recommendation transaction credit strengthens the anti-aggression of the transaction credit data. In addition, the “buyer + circle of friends” method fully uses the computing of the storage ability of the internet, and it also solves the failure problem of using a single node.
Highlights
In recent years, the e-commerce platform in China has developed dramatically and has made remarkable achievements
This paper proposes a model based on the unstructured crowd transaction network
Literature review this paper studies the credit problem of the unstructured crowd transaction network, the trust management technology of the e-commerce platform is significantly relevant to this paper
Summary
The e-commerce platform in China has developed dramatically and has made remarkable achievements. 2. Literature review this paper studies the credit problem of the unstructured crowd transaction network, the trust management technology of the e-commerce platform is significantly relevant to this paper. Li et al (2007) proposed a trust model based on similarity weighting in the P2P environment and improved the trust model based on recommendation given by Dou, which adopted a global iteration model similar to EigenRep. it is difficult for the model to resist malicious attacks, required the nodes to have good computing power and the data storage and updating program is complex, so the engineering operability is low. The lack of comprehensive consideration of transaction process factors, or the simple quantitative calculation during evaluating of transaction credit, leads to the lack of accuracy and the lack of resistance to malicious attacks for the credit evaluation model. The model is able to accurately measure the seller’s fulfillment of the transaction agreement and to resist malicious attacks in the network
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
