基于过度反应与动态网络的互联网产品扩散消费者交互行为研究

Consumer decision-making interaction behavior in Internet product diffusion based on overreaction theory and dynamic network

基于过度反应和动态网络视角,研究网络演化机制对互联网产品扩散宏观演化规律的影响。结合过度反应理论与演化博弈模型,构建个体决策概念模型,将其嵌入多智能体仿真模型,设计个体间学习模仿规则,制定多种动态网络演化机制,从而将网络动态性与扩散动态性相结合。基于Anylogic 6.5.0实现了多智能体仿真系统,模拟结果表明:在有新节点加入时,互联网产品营销活动应重点关注有大节点存在和聚类系数大的网络社区,并提高网络的连通性和聚簇程度,提升广告宣传的触达效率。在产品推广处于瓶颈期时,可采取“随机增边”的方式来增加个体连通度,并重点关注获得低收益或给予负面评价的关键节点。该研究可为互联网产品扩散的营销策略制定提供支持,并为将行为金融学理论与复杂网络应用于多智能体模型提出了基础框架。

Based on overreaction and dynamic network,the influence of network evolution mechanism on the macro-evolution law of Internet product diffusion was explored.Combining overreaction theory and evolutionary game model,the conceptual model of individual decision-making was constructed and was embedded in the multi-agent simulation model.The learning imitation rules between individuals were designed and a variety of dynamic network evolution mechanisms had been formulated to combine network dynamics with diffusion dynamics.A multi-agent simulation system was implemented using Anylogic6.5.0,and the simulation results revealed that when new nodes were added to the network,Internet product marketing activities should focus on online communities with large nodes and large clustering coefficients and expand the reach of advertising to improve the connectivity and clustering of the network;when product promotion was in a bottleneck period,"random edge increase"could be adopted to increase individual connectivity and focus on key nodes that obtain low benefits or give negative reviews.This research could make some contributions to new Internet product marketing and provide a new basic framework for combining psychology and multi-agent simulation modes.