Consumers’ Perceived Trust Transitivity Model In Social Commerce Based On Intuitionistic Fuzzy Set

Social commerce lives on social networks is the new trend of E-commerce development.But the trust crisis due to high perceived risk of online transaction and low barrier to entry has received great attention of world-wide scholars.The formation of consumers’ perceived trust has been already changed by the highly interactive social networks.Consumers establish perceived trust based on online word-of-mouth whose essence is multi-sources trust fusion.Trust transitivity model originating from computer network security focuses on information exchange between computer agents.The model always has two stages,which are trust transitivity and trust fusion.There are lots of limitation on applying traditional trust transitivity model into social commerce:(1)The primary problem is the limitation on fast processing massive recommendation into consumers’perceived trust.(2)Traditional trust transitivity model has paid little attention on the subjectivity and fuzziness of consumers’perception,because computer agents are the subject of research.The complexity of trust transitivity and multi-dimensional influnce factors leads to the difficulty of consumers’perceived trust depiction and the calculation of perceived recommendation trust in social commerce.(3)Clustering consumers’ perceived recommendation trust is an effective solution to meet the real-time requirement.But the clustering is still difficult,because online relationships between consumers,which is an important influence factor of perceived recommendation trust,cann’t be partition together with clustering of perceived recommendation trust.(4)In computer network security(especially P2P networks)research area,there is no central control mechanism to record the interaction experience.But in social commerce,consumers make transaction decision based on both recommendation and related experience.The huge different between social commerce and computer network presents new requirement to trust fusion calculation.In this paper,the framework of trust transitivity model has been extended from two stages to three stages,which are trust transitivity-clustering-trust fusion.According to the formation of consumers’ trust in social commerce,a guideline of "clustering before fusion"has been proposed to establish consumers’ perceived trust transitivity model.Therefore,the improved trust transitivity model is applicable for perceived trust real-time processing in large scale networks.Perceived trust has been depicted by Intuitionistic Fuzzy Set,and influence key factors of consumers’ perceived trust has been integrated into our model.Therefore,the trust transitivity model can be extended into social commerce.The effectiveness of our model has been tested by simulation and survey data.Main research contents are as follow:(1)Intuitionistic Fuzzy Set formation of perceived recommendation trust considering trust disposition of consumers.The key factors of perceived recommendation trust have been analyzed,consumers’ perceived trust has been depiected by Intuitionistic Fuzzy Set,and all the key factors of perceived trust have been integrated with multiple attribute decision making(MADA)to caculate the influence coefficient of recommenders.The influence coefficient has been integrated with trust disposition to adjust the intuitionistic fuzzy set of recommendation information,in order to transform the recommendation information into subjective perceived trust of consumers.(2)Relationship-oriented clustering method of consumers’ perceived recommendation trust based on spectral bisection method.Closeness network,which is exacted from social networks,has been integrated with similarity of perceived recommendation trust to improve the formation of Normal matrix.Clustering method of consumers’ perceived recommendation trust has been established by organically combining spectral bisection method based on Normal matrix and intuitionistic fuzzy set,in order to deal with the problem that consumers’ perceived trust and social relationships can’t be both reflected neither in clustering methods nor in spectral bisection method.(3)Multi-sources trust fusion model considering consumers’ expierence.First,herd behavior has been considered to fusion all the clusters of perceived recommendation trust to reflect consumers’ psychological character of "the minority is subordinate to the majority".And the fusion method of intuitionistic fuzzy set has been improved by subjevtive logic method to deal with the dramatically decrease of nonmembership of intuitionistic fuzzy sets addition.Second,the Case-based Decision-making Theory(CBDT)has been extended and integrated with intuitionistic fuzzy sets theory to calaulate consumers’ perceived empirical trust.Third,multi-sources trust fusion model has been established by introducing anchoring effect into fuision method of perceived recommendation trust and perceived empirical trust.In this paper,the trust transitivity model,which has two stages,has been extended to three stages:trust transitivity-clustering-trust fusion.Consumers’ perceived trust has been divided into perceived recommendation trust and perceived empirical trust,and several methods such as intuitionistic fuzzy sets,spectral bisection method,MADA,CBDT,and Subjective Logic has been merged to establish consumers’ perceived trust transitivity model in social commerce.The improved model can express subjectivity and fuzziness of consumers’perceived trust,provide solution to massive recommendation information processing in large-scale social networks,contribute to trust transitivity research system improvement,and provide new way to forecast and analyze consumers’ perceived trust in social commerce.Our model can be applied as trust-based recommendation system kernel in social commerce,and assist social commerce platform in making marketing strategy.