基于社会认知理论的电子商务动态信任网络模型研究

In E-Commerce environment, the configuration and evaluation of trust relationships among transaction entities is one of the efficient technical means that controls the transaction risk. However, both the relationship attributes of the recommendation entities and the relationship between trust and risk evaluation are not taken into account in the existing researches, both the dependability and the practicality of the existing trust computing models can not meet the practical application need. .Therefore, in order to explore the essential features of the trust relationships under e-commerce environment,the attributes and the trust relationships for transaction entities are studied and defined in terms of social cognitive theory and method. Also, this proposal presents a formal description model based on the trust network. Considering to "word of mouth" propagation laws, this proposal proposes configuration and optimization algorithms for the trust network which can effectively simplify the complex network relationship, it is hopeful that the dynamic deduction law of the trust is discovered for transaction entities.On Basis of this, the proposal puts forward a multi-dimensional dynamic trust algorithm and a trust risk evaluation algorithm so as to construct a control system of the trust risk for e-commerce transactions. Finally, several simulation expereiments are carried out in order to verify the scientificity and effectiveness of the proposed trust computing algorithm and the trust risk evaluation algorithm, and build the foundation for the dynamic trust network theory system of e-commerce..The research is of profound theoretical and practical significance for the sound development of e-commerce. In addition, it is of instructional and referenced purpose for researchers to study the trust relationships among entities of virtual communities (e.g. net friends, exclusive circle members, blog friends, etc.) in Internet.

建立和评价交易实体间的信任关系是控制电子商务交易风险的有效技术手段之一。然而，已有的信任计算模型并没考虑实体(推荐节点)间的关系属性以及信任与风险关系的测度，可信性和实用性不能满足实际需求。.基于此，本项目拟将社会认知理论和方法引入到电子商务中来，着重研究交易实体及实体间的关系属性，构建电子商务信任网络形式化描述模型；应用社会认知中的口碑传播规律，研究并提出信任网络构造和优化算法，有望揭示电子商务实体间信任的动态演绎规律；并以此为基础，构建多维度动态信任计算模型和信任与风险关系测度模型，以期建立电子商务交易风险防范体系；通过实验仿真，验证信任计算和信任与风险关系测度模型的科学性和有效性，为构造电子商务动态信任网络理论体系奠定基础。.其研究对推动电子商务健康发展具有重要理论意义和实用价值，同时对网络环境下虚拟社区实体间(如网友、圈友、博友等)的信任关系研究具有参考借鉴作用。

建立和评价交易实体间的信任关系是控制电子商务交易风险的有效技术手段之一。然而，已有的信任计算模型并没考虑实体(推荐节点)间的关系属性以及信任与风险关系的测度，可信性和实用性不能满足实际需求。. 基于此，本项目应用社会认知理论和方法，着重研究了电子商务环境中交易实体及实体间的关系属性，定义了信任网络以及信任网络中涉及到的一些基本概念，提出了一种基于有向图的电子商务信任网络形式化描述方法；借助于(a-r)-energy模型，定义了信任网络中节点的公共影响力权重、节点间的信任距离、节点引力势能和斥力势能的概念，提出了一种基于力导引模型的信任网络可视化算法和可视化工具；应用社会认知中的口碑传播规律，研究了信任在信任网络中的传播规律和信任路径融合算法，揭示了信任的时间敏感性、不对称性、可传递性和可选择性等属性，并以此为基础，构建了多维度动态信任计算模型和信任与风险关系测度模型；通过实验仿真，验证了信任计算和信任与风险关系测度模型的科学性和有效性，为构造电子商务动态信任网络理论体系奠定了基础。