基于贝叶斯网络的化学恐怖袭击后果评估方法研究

Research on Evaluation Method of Chemical Terrorism Attack Consequence Based on Bayesian Network

作为一种非常规形式的恐怖袭击,化学恐怖袭击造成的人员伤亡和社会影响往往极为严重。通过梳理恐怖袭击发生的主要脉络,利用K2算法构建贝叶斯网络结构,运用EM算法进行参数学习,得到贝叶斯网络风险评估模型。实验结果表明,利用建立的贝叶斯网络模型,得到人员伤亡等级分类准确率为0.75,该模型可用于推理分析和实例预测。通过推理分析发现,相比于爆炸类化学武器,中毒类化学武器被使用的概率更高,应该受到更多关注。利用发生在阿富汗和伊朗的4起化学恐怖袭击事件进行风险评估模型的验证。结果表明,模型评估结果和事件发生后果相符。研究结果预期可以为公安机关对化学恐怖袭击的预防提供决策支持。

As a type of unconventional terrorism,chemical terrorist attack often causes large number of casualties and extremely serious social impact.In this study,main process of terrorist attacks is described,and based on that Bayesian network structure is also constructed by K2 algorithm.Moreover,EM algorithm is used for parameter learning to establish the Bayesian network risk assessment model.The results show that the classification accuracy of casualty levels of the established Bayesian network model is 0.75.The model can be used for reasoning analysis and instance prediction.Through reasoning analysis,compared with explosive chemical weapons,poisoning chemical weapons are more likely to be used,which should receive greater attention.Finally,four real cases of chemical terrorist attacks in Afghanistan and Iran are used to validate the risk assessment model,showing that assessments are consistent with the consequences of the real event.The findings of this paper are expected to provide decision support to the police for chemical attacking prevention.