关键词: 过度反应, 消费者决策, 多智能体模拟, 动态网络, 互动行为

Abstract: 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.

Key words: overreaction theory, consumer decision-making, multi-agent simulation, dynamic network, interaction