Trust Model Based on Bayesian Network in E-commerce

Trust is a key constraining factor in the development of electronic commerce. The parties must establish a certain trust model between each other before the beginning of transaction. This paper analyses the advantages and deficiencies of existing models, and presents a trust model for E-commerce based on Bayesian network. The model is based on Bayesian network and conditional probability formula, introducing conditional probability table, fully considered the impact of major factors which have relationship to formal e-commerce transactions, raising the transaction success rate to a certain extent. Meanwhile, this model introduces the time decay factor and the incentive factor, reducing the generating probability of malicious node.