摘要
Web病毒式营销中,客户间的信任关系类似于传统商务环境中实体间的社会关系,信任网络类似于传统商务环境中复杂的社会网络,客户间的信任度量受客户认知水平、交易金额、交易时间等因素影响,难以给出准确的量化.为探明这种信任网络的本质内涵以及降低信任关系的复杂性,结合电子商务的认知理论和方法,详细分析和定义了实体属性以及实体关系.给出一种基于客户特征数据和交易数据的客户信任网络模型CNTM(Customer Network of Trust Model),提出三种不同应用环境下的信任网络构建策略COS(Customer Occupation based Strategy),CIS(Customer Interest based Strategy)和COIS(Customer Occupation and Interest based Strategy),并分别给出信任度计算方法.与已有研究相比,提高了信任网络的覆盖率,保证了信任网络的信任强度并有效的改善了冷启动问题.
In the process of Web viral marketing, trust relations among customers are similar to the social relations among entities in the traditional business environment; and trust network is similar to the traditional complex social network. The trust evaluation be- tween customers is hard to measure due to factors such as customer＇s knowledge level, transaction scale, trading time and so on. For the purpose of understanding the essential connotation of trust network and reduce the complexity of trust relations, this paper pro- posed a Customer Network of Trust Model { CNTM ） based on customer features and transaction data. In the model, entity properties and entity relationships are analyzed and defined in accordance with the knowledge theory and approaches of e-commerce. Three dif- ferent strategies of trust network construction, Customer Occupation based Strategy { COS }, Customer Interest based Strategy { CIS } and Customer Occupation and Interest based Strategy （ COIS }, are introduced in this paper; and the method of trust computing in each strategy is given. Compared with other studies, the coverage rate of trust network is improved, the trust strength of trust network is guaranteed and the cold start problem is improved.
出处
《小型微型计算机系统》
CSCD
北大核心
2013年第9期2075-2080,共6页
Journal of Chinese Computer Systems
基金
辽宁省自然科学基金项目(201102172)资助
