基于空间数据分析的长江公共安全预测模型构建

Construction of the Yangtze River Public Security Prediction Model Based on Exploratory Spatial Data Analysis

[目的/意义]长江流域安全关系着国家政治经济的安全稳定,对其公共安全事件的预测效度显得尤为重要,但目前尚缺乏较好的预测技术与方法。[方法/过程]探索性空间数据分析是风险评估和警务预测中的一个热点技术,可以提炼数据并应用于事件描述、聚类关联以及评估预测等。以警用地理信息系统为依托,构建犯罪热点分析、邻近重复分析和风险地形建模混合模型,综合运用ArcGIS、RTMDx等空间软件进行计量、分析、建模,对长江流域苏锡段公共安全事件进行预测。[结果/结论]运用探索性空间数据分析,可综合研判可能影响公共安全事件发生、发展、转移和变化的相关因素,对未来潜在公共安全事件的分布、结构和趋势等作出评估与预测。研究结果发现:犯罪热点分析、邻近重复分析、风险地形建模混合模型具有较好的预测效度,模型准确预测了91%的长江公共安全事件。

[Purpose/Significance]The security of the Yangtze River Basin is closely related to the security and stability of the country's politics and economy.It is particularly important to predict the validity of its public security events,but there is still a lack of better prediction technology and methods.[Method/Process]Exploratory spatial data analysis is a hot technology in risk assessment and predictive policing.It can extract data and apply it to event description,cluster correlation,evaluation and prediction,etc.Based on the police geographic information system,a hybrid model of crime hot spot analysis,neighborhood repeat analysis and risk terrain modeling is constructed,and the public security events in the Suzhou-Wuxi section of the Yangtze River Basin are predicted by comprehensively using ArcGIS,RTMDx and other spatial software for measurement,analysis and modeling.[Result/Conclusion]Using exploratory spatial data analysis to comprehensively study and judge the relevant factors that may affect the occurrence,development,transfer and change of public security events,the authors make assessment and prediction of the distribution structure and trend of potential public security events in the future.It is found that the mixed model of crime hot spot analysis,proximity repeat analysis and risk terrain modeling has good prediction validity,and the model accurately predicts 91%of public security events.