美国犯罪情报预测分析技术的特点——基于兰德报告《预测警务》的视角

The Characteristics of the Criminal Intelligence Predictive Analytical Technology in the United States: an Analysis from Rand's Report

[目的/意义]犯罪情报预测分析技术稀缺,一直是制约我国开展大数据犯罪情报分析的瓶颈问题。借鉴美国先进的预测分析方法和技术,对提升我国犯罪情报预测分析的水平十分有益。[方法/过程]以美国兰德公司发布的《预测警务》报告为视角,解析了美国犯罪情报预测分析的方法、层级和技术指标。[结果/结论]研究发现,基于大数据的犯罪情报预测分析的技术呈现出四个特点:一是美国犯罪情报预测分析的技术路线是"数据统计+数据挖掘+犯罪制图";二是美国犯罪情报预测分析的核心是预测犯罪趋势;三是美国犯罪情报预测分析重点解读的是犯罪要素的关联关系;四是美国犯罪情报预测分析的主角是专业的犯罪分析师。

[Purpose/Significance] The scarcity of criminal intelligence predictive analytical technologies is always the bottleneck that re-stricts the big data analysis potential development for criminal intelligence analysis in China. Studying and learning from the methods and technologies of predictive analysis in the U. S. is beneficial to improve the situation. [ Method/Process] This article analyzes the method, layers and technical indicators of the criminal intelligence predictive analytical technology based on the perspective of Predictive Policing published by the Rand Corporation. [ Result/Conclusion] According to the study, the technology of the predictive analysis on criminal in-telligence based on big data highlights four characteristics：the technical route of the criminal intelligence predictive analysis is＂｜statistics＋ data mining ＋crime mapping＂;the core of the criminal intelligence predictive analysis is to predict crime trends;the focus of the crime intelligence predictive analysis is on the interpretation of the relationships between the elements of the offence;the protagonist of the crimi-nal intelligence predictive analysis is a professional crime analysts.