Research On E-Commerce Reputation Management Model Based On Fuzzy Theory

Trust problem of E-commerce has been increasingly a focus of people's attention. Studies by researchers show that e-commerce reputation of enterprises has positive impact to enhance online customer trust, and could bring continuing benefits for business. Business competition based on reputation management becomes the new strategy for business development in e-market. The strategy distinguishes from the theory of leveraging legal mechanisms to regulate electronic markets so as to boost market confidence. In the open network environment, the reputation relationships among entities are dynamic. The e-business reputation management involves many different elements: the collaboration among entities, collection of the information to assess the reputation, and the standards for reputation assessment. As e-business reputation is subjective, ambiguous, dynamic, which could not be accurately described and verified, it is necessary to seek new solutions.The contributions of this dissertation are as follows:1) A gaming model of e-business reputation management is presented. The model is established both in the incomplete and complete information conditions. Under different conditions, it studies the mixed cost-efficiency strategies and analyzes the choice solutions of reputation model between business and consumer. It explains that in the long-term transaction relationship, businesses will strive to establish image and maintain reputation to have long-term cooperation revenue. The model also discusses the countermeasures of discount rate, disciplinary measures and mechanisms to enhance reputation.2) An e-commerce subjective reputation management model based on fuzzy is presented. The model gives the definition of e-commerce in the network environment, and analyzes the reputation attributes of asymmetry, transferability, dynamic changes, and the over time attenuation. Reputation vector is set up by the use of ambiguous language; the formal reasoning of reputation vector is conducted; algorithm of fuzzy connection and joint operations is developed in order to study the mechanism of e-business reputation formation. The model also proposes a fuzzy reputation clustering method of dynamic type definition mechanism to cluster entities with different reputation levels.3) The analysis of e-commerce initial reputation based on trust is presented. It explains that consideration factors for business which has just entered the market with no historical trading background. The concept of initial reputation is given. Fuzzy reasoning is carried to analyze the self-suggestion and recommendation trust to study the constructing of initial reputation and the contribution of trust factors in e-commerce system. Through literature review, the design factors are found for consideration of e-commerce reputation evaluation.4) A dual-fuzzy comprehensive judgment model of e-business reputation evaluation is presented. The first "ambiguity" refers to the set up of a multi-level fuzzy evaluation model including reputation membership function construction, first-level evaluation, second-level evaluation, and multi-level evaluation. The second "ambiguity" refers to the fuzzy analysis of the importance weight factor by using the improved analytic hierarchy process method to calculate the values of importance of factors. That is the matrix of importance factor can be set up first, then determine the correlation orders by using gray incidence analysis, and use AHP method to determine the weight of each factor.