Drug-related crime risk assessment and predictive policing based on risk terrain modeling
ZHANG Ning, WANG Dawei*,
School of Criminology, People's Public Security University of China, Beijing 100038, China
Changjiang River Administration of Navigational Affairs in Ministry of Transport of the People's Republic of China's Technology Project, No.201710012;
Crime is the product of a certain time and space. Research on crime cannot be separated from temporal and spatial analyses, as well as social, geographical, ecological, environmental, and other factors that generate crime. Risk terrain modeling technology was developed by American scholars for spatial risk assessment and predictive policing. It has been independently proven and tested in over 45 countries across six continents around the world and 35 states in the United States. It has been widely used in many fields such as predictive policing, homeland security, traffic accidents, public health, child abuse, environmental pollution, and urban development. It has achieved remarkable results in the crime research area of drug, arson, explosion, rape, robbery, and theft. This study adopted crime hotspot analysis and risk terrain modeling to analyze the risk factors, spatial blind spots, and risk terrain of narcotics crimes in 2015 in N City of the Yangtze River Delta region, explored the mechanism and evolution of drug crimes, and made a prediction on N City 2016 drug crime trend. The results show that N City drug crime presents obvious crime hotspots and crime cold spots. Rental housing, hotels, railway stations, banks, parking lots, entertainment venues, urban expressways, and Internet cafes are drug risk factors in the city. Risk terrain modeling is effective in predicting drug crimes. The narcotics departments of public security organs should put more police and energy to gradually limit and eliminate the hotspots that generate, attract, and promote crime.
毒品犯罪起因复杂、蔓延迅速、后果严重、影响广泛,已成为各国政府及国际组织高度重视的重大社会问题。据联合国毒品与犯罪办公室2015年报告显示,全球每年约有2.5亿吸毒者,其中有20.7万例与吸毒有关的死亡案例(United Nations Office on Drugs and Crime, 2016)。根据中国国家禁毒委员会办公室发布的《2017中国禁毒报告》,2016年共破获毒品刑事案件14万起,查处有吸毒行为的人员100.6万名。报告指出,截至2016年底,中国有吸毒人员250.5万名,吸毒人员低龄化特征明显。由于各类新型毒品相对容易取得,加上年轻人求新、求异、求奇等因素及虚幻世界诱惑,毒品防控工作面临着新挑战。毒品犯罪常与暴力、卖淫、赌博、艾滋病、贫困、青少年、家庭破裂等紧密相连。Gorman等(2004)通过最小二乘法研究发现,毒品犯罪密度可以解释全部暴力犯罪71%的变异。Caeti(1999)在1994-1996年与休斯顿警察局合作针对毒品犯罪进行了一项实验研究,透过官方资料及实地观察结果选定7个犯罪率较高的犯罪热区,并采用提高见警率、犯罪零容忍政策以及问题导向警务,研究结果发现,实验组随着毒品犯罪率的下降,抢夺、盗窃犯罪均较控制组显著下降。此外,毒品犯罪作为无被害人犯罪,还与有组织犯罪、黑社会犯罪相互交织、相互渗透、互为促进。全球黑社会组织与毒品犯罪集团正以不同形式联手,成为威胁全球的不安定因素,对人类的生存与发展构成严重挑战。
许多研究证实,犯罪地点在城市空间上并非随机分布,而呈现一定集聚性(Chainey et al, 2005; 刘建宏, 2012; Santos et al, 2012)。如能阻止这些热点地区的犯罪活动,就可有效减少社会的总体犯罪活动。美国泽西市毒品市场分析计划、费城摧毁街角毒品犯罪计划均取得较好的效果。泽西市犯罪转移和扩散干预研究项目实施后,犯罪热点的贩毒活动减少了58%(Braga et al, 2014)。随着地理信息系统(Geographic Information System, GIS)的迅速普及与发展,国外学者开始以犯罪时空为视角,分析犯罪的分布、集聚、转移、变化等特征和规律,以及犯罪的生成机理和演化扩散规律,并取得一定的成效。在中国,随着公安部“金盾工程”二期建设的顺利推进,中国各地公安机关相继开发应用警用地理信息系统(Police Geographic Information System, PGIS)。但从目前的实际运用来看,各地各部门大多仍停留在简单的储存、展示、查询、制表等功能,地理信息系统强大的空间分析功能没有得到充分挖掘和应用。如何将地理信息系统与大数据、机器学习、人工智能、警务预测充分融合,必将成为进一步深化警用地理信息系统应用的重要课题。伴随着科技进步与物联网、云计算、社群媒体、智能手机及大数据分析逐渐盛行,如何透过风险地形建模和警务预测系统,对大量的数据进行统计、比对、解析,以得出较客观化、科学化、精准化的分析结果,对日益加剧的毒品犯罪进行有效的预测、预警、预控,将是当前乃至今后很长一个时期非常重要的研究课题之一。当前中国毒品流通渠道泛滥,若能掌握毒品犯罪空间分布形态并加以分析,针对易发生施用或交易毒品热点,探究其区域特性,并采取相应的防控措施,对于打击和抑制日益加剧的毒品犯罪具有重要作用。
本文将影响N市2015年毒品犯罪风险性因素分为4大类25个小类,其中4大类分别为：生产基础设施、商业基础设施、社会基础设施和社会控制机构。所有的空间数据统一投影并进行空间配准,通过位置、地点、街长、面积等空间参数的量算,将N市划分为107042个网格。回归过程中确定了N市毒品犯罪最佳拟合风险地形模型是一个负二项Ⅱ模型,保留了初始25个风险项因素中的8个变量。根据贝叶斯信息准则(Bayesian Information Criterion, BIC),在双向逐步回归的过程中,利用期望值和修正概率做出最优决策,最后构建最优模型,BIS得分为8571.7分。表1是N市毒品犯罪的最优风险地形模型,其中包括8个风险因素,以及相对应的空间影响、系数和相对风险值。最具影响力的风险因素是出租屋,相对风险值为26.08(意味着其风险是相对风险值为1的因素的26.08倍);其次酒店相对风险值为2.45;此外,车站、ATM机周边、停车场、娱乐场所、城市快速路、网吧的相对风险值分别为2.36、2.16、1.89、1.71、1.48和1.39。
Optimized risk terrain modeling (RTM) model specifications
By considering the influencing factors such as time and distance, crime rate, population, police, geography and environment, victims鈥 occupation, etc, the authors use mathematical modeling to establish the evaluation function of crime probability in the research area, calculate the probability of crime, and then combine it with GIS-related technology to get the most likely crime areas, namely, the geographic portrait of crime. The authors conduct method validation and analysis through examples. This new probabilistic crime model could provide geospatial data for detecting serial criminal cases and narrow down the scope of police surveillance. This new investigative technique with high precision is suitable for various geographic regions and helpful for detecting serial criminal cases.
[XiaoH, Du YH, Xu JZ, et al.2013. Integration of the GIS with criminal probability model and its application[J]. , 49(6): 1025-1030.]
[Ye BC.2016. A preliminary study on the application of the third and level 4 drug related crime in juvenile and adult: A case study of Taipei city[J]. , 17(3): 57-88.]
Barnum JD, Campbell WL, TroccioS, et al.2016. Examining the environmental characteristics of drug dealing locations[J]. ,6: 1-26.
Abstract Illicit drug dealers who operate in open-air markets must access customers in the face of risks posed by law enforcement, customers, and competitors. However, researchers purport that the environmental characteristics of certain locations may allow dealers to balance these competing demands. This research utilizes risk terrain modeling to identify the environmental characteristics of drug dealing locations in Chicago, Illinois, and compares these characteristics across different types of drugs. Results show that a number of place features increase the risk of drug dealing. Although many of these features are common to locations for all types of drug dealing, there are some variations. The findings provide insight into the environmental context of drug market locations, which has implications for researchers and practitioners.
Braga AA, Papachristos AV, Hureau DW.2014. The effects of hot spots policing on crime: An updated systematic review and meta-analysis[J]. , 31(4): 633-663.
In recent years, crime scholars and practitioners have pointed to the potential benefits of focusing police crime prevention efforts on crime places. Research suggests that there is significant clustering of crime in small places or ot spots.A number of researchers have argued that crime problems can be reduced more efficiently if police officers focused their attention to these deviant places. In this article, we update and improve upon a previously completed Campbell Collaboration systematic review of the effects of hot spots policing and crime. Meta-analyses were used to determine the size, direction, and statistical significance of the overall impact of hot spots policing strategies on crime. The results of our research suggests that hot spots policing generates small but noteworthy crime reductions, and these crime control benefits diffuse into areas immediately surrounding targeted crime hot spots. Our analyses find that problem-oriented policing interventions generate larger mean effect sizes when compared to interventions that simply increase levels of traditional police actions in crime hot spots. We also find that only a small number of studies examine the impacts of hot spots policing on police-community relations. The extant research on this topic, however, suggests that community members have positive reactions to these focused policing actions.
Caplan JM, Kennedy LW.2016. [M]. Berkeley, CA: University of California Press.
ChaineyS, RatcliffeJ.2005. GIS and crime mapping[M].
Gaziarifoglu Y, Kennedy L W. 2014. Applying risk terrain modeling to street robberies in Newark, NJ[J/OL]//. 2014-11-25 [2018-08-21]. .
Gorman DM.2004. Alcohol outlet density and violence: a geospatial analysis[J]. , 4(39): 369-375.
Abstract AIMS: To examine the relationship between alcohol outlet density and violent crime controlling for neighbourhood sociostructural characteristics and the effects of spatially autocorrelated error. DESIGN: The sample for this ecologic study comprised 188 census tracts from the City of Austin, Texas and 263 tracts from the City of San Antonio, Texas. Data pertaining to neighbourhood social structure, alcohol density and violent crime were collected from archival sources, and analysed using bivariate, multivariate and geospatial analyses. RESULTS: Using ordinary least squares analysis, the neighbourhood sociostructural covariates explained close to 59% of the variability in violent crime rates in Austin and close to 39% in San Antonio. Adding alcohol outlet density in the target and adjacent census tracts improved the explanatory power of both models. Alcohol outlet density in the target census tract remained a significant predictor of violent crime rates in both cities when the effects of autocorrelated error were controlled for. In Austin, the effects of alcohol outlet density in the adjacent census tracts also remained significant. The final model explains 71% of the variance in violent crime in Austin and 56% in San Antonio. CONCLUSIONS: The findings show a clear association between alcohol outlet density and violence, and suggest that the issues of alcohol availability and access are fundamental to the prevention of alcohol-related problems within communities.
HollernanD, GaleR.2013. An application of risk terrain modeling to residential burglary[J]. , 15: 1-9.
ABSTRACTUsing dispatch data from the Lawrence Township Police Department (LTPD), this studyexamines the extent to which indicators of "risk," as identified by the prior research and routine activities theory, correlate with the geospatial distribution of burglary via the application of risk terrain modeling (RTM). Risk is operationalized in this study by the geospatial concentration of bus stops, calls for suspicious vehicles, and calls for suspicious persons in Lawrence Township. Results indicate that those areas in Lawrence Township with highest concentration of bus stops, calls for suspicious vehicles and calls for suspicious persons also have the highest concentration of burglary events.
Santos RB.2012. Crime analysis with crime mapping[M].
United Nations Office on Drugs and Crime. 2016. UNODC annual report 2015[EB/OL].
WeisburdD, Mazerolle LG.2000. Crime and disorder in drug hot spots: Implications for theory and practice in policing[J]. , 3(3): 331-349.