社群用户信息决策中级联效应判识的修正

Correction on Identification of Cascade Effect in Information Decision of Community Users

针对传统级联效应判识模型中存在的问题,分析了基于马尔可夫链对级联效应判识模型进行修正的可行性。通过改进后的马尔可夫链对信息级联的判识做出了修正,降低了后验概率计算的繁杂程度,同时突出了新近事件在决策中的权重。在提出了社群中用户所处决策位序与级联效应发生的相关性的同时,从数理层面揭示了社群中信息级联的脆弱性与级联自我校正的内在规律。最后基于模型仿真的结果对社群用户信息决策中的级联效应进行了讨论。

For problems in the traditional model of identification of cascading effect, this article analyzes the feasibility of improvement model of identification of cascading effect based on Markov chain. The iden- tification of cascade information is corrected through improved Markov chain. The optimizing reduces the complexity of posterior probability calculation, while highlighting the weight of recent events in the deci- sion-making. It announces the intrinsic law both of fragility and self-correction from mathematical view, while the relevance between sequence number and occurring of cascade effect is proposed. Finally, a＇ sim- ple discussion on the cascade effect in information decision of community users is presented based on model simulation.