E-commerce Based On Similarity And Network Optmization Research On Trust Technology

Owing to the rapid development of network technology and increasingly opening network environment,electronic trading activities and related services has been developed rapidly,giving people's lives great convenience.But the rapid development of e-commerce also gradually showing some drawbacks,such as inconsistent description with the actual goods,logistics speed mismatch and malicious buyers or sellers,etc..These problems are attributed to e-commerce's trust mechanism problem,it has become the biggest obstacle to development of electronic commerce.To solve this problem,scholars have put forward by unremitting efforts a variety of trust models,mainly including how to build a safe,effective and scientific decision-making mechanism of the transaction in order to ensure that consumers get the maximum benefit protection.This paper aims completes the following works aims at e-commerce trust model:(1)It analyses the research status of e-commerce trust model,defines related concepts,and some classical e-commerce trust model are disused briefly.(2)It propose a trust model based on the similarity of the buyer and transaction services,introducing the concept of sellers of private trust and sellers trust of the public trust.Firstly,the buyer and the recommender's similarity degree are selected to add some creditors to the list of trusted recommender.Secondly,based on the list of trusted recommender,calculate seller's private trust and public trust.Finally,we compute the seller's comprehensive trust,and choose the trustworthy seller for the buyer by the expected value function.Simulation results show that the model can improve the accuracy of trust calculation,and help buyers select trusted sellers to improve transaction success rate.Also this model has a strong anti-attack ability.(3)It proposes a calculation model of network optimization and trust network based on trust network.Firstly,3 rules of network optimization are put forward,which is used to optimize the complex trust network;secondly,based on the original recommendation path algorithm,calculation method of trust value based on path depth is proposed,including the two aspects: the single path recommendation andparallel paths but independent recommendation.Finally,comprehensive trust value of the source node to the target node is calculated from multi-path of trust.Instance analysis and simulation experiments show that the model has better advantage in the trust value calculation of multi-path recommendation,it has a higher accuracy the source node to the target node trust calculation.Because of the trust network is optimized in advance,so the model can effectively reduce the time complexity,and it also has advantages in resisting malicious user nodes.