Abstract: In the context of social media, studying the impact of trust relationship on the process of online word-of-mouth (oWOM) communication can help companies formulate reasonable marketing strategies to control oWOM communication. First, based on the SIR epidemic model, a continuous time Markov oWOM communication model based on trust relationship is constructed, where social reinforcement and consumers’ interests shifting phenomenon are considered. The degree of trust between consumers is characterized by the strength of their social ties. Then, the effects of trust relationship on oWOM communication in networks with different community structures are analyzed using numerical experiments. Experimental results show that the improvement of trust can significantly promote the diffusion of oWOM. When the communities overlap, the promotion effect of community trust is enhanced, while the role of social trust is weakened. In addition, compared with the promotion effect of community trust on oWOM communication, the effect of trust among strangers is stronger, especially when the community structure of the network is not strong. In this case, influencers can more effectively control the oWOM communication process.

Key words: online word-of-mouth；SIR (susceptible, infective, removal) model；trust；community structure；numerical analysis

摘要： 在社会化媒体环境下，研究信任关系对在线口碑传播过程的影响，有助于企业制定合理的营销策略控制口碑传播。以社会关系强度刻画消费者之间的信任度，考虑社会强化现象及消费者兴趣转移现象；以SIR传染病模型为基础，构建基于信任关系的连续时间马尔科夫在线口碑传播模型。通过数值实验，分析了不同网络结构下信任关系对在线口碑传播的影响。实验结果显示：信任度的提升可显著促进在线口碑传播，当社区存在交叠时，社区信任度的促进作用增强，而社会信任度作用则减弱；另外，相比于社区内部信任，提升陌生人之间的信任促进作用更强，尤其是当网络社区结构一般或较弱时，且此时高影响力用户可在控制口碑传播过程中发挥更大作用。

关键词: 在线口碑, SIR模型, 信任, 社区结构, 数值分析