犯罪案件时空热点分析研究

【作者】 王帅；

【导师】 李小娟； 王法辉；

【作者基本信息】 首都师范大学 ， 地图学与地理信息系统， 2012， 硕士

【摘要】 犯罪案件时空热点分析始于1829年,经历了开创、研究与发展、应用和商业化、客户端应用程序等一系列发展。作为一个警务人员的辅助工具,为犯罪预防、行动有效性评估、各层级平台间交流、警力规划、和检验犯罪理论方面提供有力支持。从空间的角度分析犯罪事件可以直观地反映分布特点、趋势,便于理解与时间、地点及周围事物的相互关系。犯罪空间分析发展至今,汇集了大量跨领域学者,共同努力形成了较为成熟的算法与流程,并不断探索创新。目前,我国的犯罪空间分析中存在着重现象不重本质,重空间不重时间,重宏观不重微观等现象,针对以上不足,本文以江苏省某市为研究区域,选取2009年10月至2011年8月的“盗窃电动自行车”、“盗窃民宅”和“两抢”(抢劫和抢夺)案件数据,利用经典的地学统计模型及时空分析的新方法,对三类案件进行了完整的热点时空分析工作,主要包括以下几个方而：(1)提出一套完整的时空犯罪热点识别流程,包括犯罪空间分布模式识别、犯罪热点可视化分析及热点时间预测。(2)研究犯罪时空热点识别与可视化方法,从时间和空间、基于点和基于面的角度确定了三类案件各自的热点分布。计算结果清晰地展示了研究区域的犯罪热点分布状况和与敏感地区的相关关系。(3)分析犯罪成因,讨论热点内部犯罪与商场、学校等敏感场所与案件的距离关系。并借助社会环境经济数据如人口、面积、人均GDP、就业率、医院、学校数量等,从人与社会环境关系角度分析热点成因。证明了不同案件对相同的敏感场所的空间相关关系各不相同,案件发生主要受人口、面积和人均GDP的影响。(4)将求和自回归滑动平均模型(ARIMA)引入犯罪案件时间热点分析,利用现有的案件分布现状预测未来的犯罪热点,以达到预防犯罪作用。结果显示预测值与实际较为相符。基于上述的研究证明了论文整个流程的严谨性和可行性,丰富了当前我国的警用地理信息系统建设研究的案例。最后对本文主要结论进行陈述并展望犯罪案件时空热点的发展趋势。更多还原

【Abstract】 Time-space crime analysis began in1829, has experienced a series of development, the foundation, research and development, application and commercialization, the client application. As an auxiliary tool for police officers, it provides strong support on crime prevention, action effectiveness evaluation, exchanges of views between each level of the platform of police, and criminal theory test. From the perspective of spatial analysis of crime can intuitively reflect the distribution characteristics, trend, and facilitating understanding the relationship between time, spatial and surrounding things. Crime analysis of spatial development bringing together a large number of interdisciplinary scholars, and formed a relatively mature algorithms and processes, and continuously explore innovative method.At present, there are some problems in our country’s crime spatial analysis, lay too much stress on spatial analysis but not on temporal analysis, and on macroscopical phenomenon not on microcosmic phenomenon, In view of the above,this paper selects a city in Jiangsu Province as a study area. The research datas are robbery (Rob and plunder), Burglary and electric bicycle theft cases from2009October to2011August., this paper mainly discusses four issues:(1) Summarized a set of complete spatial-temporal crime hot spot identification process, from the perspective of time and spatial for popular identify method and analysis respectively, identifies three categories of cases the respective distribution of hotspots.(2) Focus on the sensitive areas inside the hot spots. Aiming at hot spots to discuss distance relation between shopping malls, schools, other sensitive places and crime cases. The actual results show that cases of various types of hot spots were identified and displayed the distribution clearly. Different cases show different correlation in the same sensitive places.(3) Reason analysis of crime hot spots. Then by means of the social environment and economic data such as population, area, per capita GDP, employment rate, the number of hospitals and schools, research the causes of hot spots from the perspective of relationship between man and society environment. Prove the cause of crime is the main population, area and per capita GDP.(4) Predicting hotspots throuth the distribution of the cases. this paper predicts crime hotspots to prevention crime. And the outcome of prediction case is more correspond with practice.So that the entire process of research is rigorous and feasibility, enriches the construction of police geographical information system case study our country.更多还原