基于时空临近重复效应的犯罪热点特征及成因分析——以北京市抢劫案件为例
郭雅琦(1996— ),女,河南郑州人,硕士生,主要从事犯罪地理、文本挖掘方面的研究。E-mail:hnpyagyq@163.com |
收稿日期: 2019-10-12
要求修回日期: 2020-02-25
网络出版日期: 2020-07-28
基金资助
北京市自然科学基金项目(9192022)
社会安全风险感知与防控大数据应用国家工程实验室主任基金项目()
中国人民公安大学2019年拔尖人才培养专项资助硕士研究生科研创新项目(2019ssky002)
版权
Formation and feature analyses of crime hotspots using near repeat principle: A case study of robbery in Beijing
Received date: 2019-10-12
Request revised date: 2020-02-25
Online published: 2020-07-28
Supported by
National Natural Science Foundation of Beijing(9192022)
National Engineering Laboratory Director Fund for Social Security Risk Perception and Prevention and Control of Large Data Applications()
Top Talents Training Specialized Subsidy for Scientific Research and Innovation Projects of Master's Graduates of People's Public Security of China in 2019(2019ssky002)
Copyright
时空临近重复效应是犯罪活动的一种重要时空特征。为深入研究犯罪热点的特征及其形成原因,论文以北京市内城六区2012—2014年抢劫案件为例,通过核密度估计、时空临近重复计算及定义时空临近重复案件链等方法分析了犯罪热点的案件构成,并从犯罪人因素和环境因素等方面对犯罪热点内的案件特征结构进行了分析。结果表明:北京市内城六区的抢劫案件存在有“a”“b”“c”三个主要的空间热点,并且热点内的大部分案件均具有显著的时空临近重复效应;其中热点“a”位于双井、劲松一带,热点“c”位于南四环大红门桥一带,且2个热点内案件的犯罪人特征在一致性程度上高于环境类特征,表明热点的形成源于犯罪人在热点区域内重复作案的可能性较大;而热点“b”位于东南三环的分钟寺地区,热点内案件的环境类特征在一致性程度上高于犯罪人特征,表明该热点的形成为不同犯罪人在热点区域内集中作案的可能性较高。研究对警务部门开展针对性的犯罪打击和防控有一定的支撑作用。
郭雅琦 , 陈鹏 . 基于时空临近重复效应的犯罪热点特征及成因分析——以北京市抢劫案件为例[J]. 地理科学进展, 2020 , 39(5) : 804 -814 . DOI: 10.18306/dlkxjz.2020.05.009
As a significant spatiotemporal characteristic of crimes, repeat and near repeat pattern has received much interest in criminology research. The purpose of this study was to explore the formation process and features of crime hotspots by using near repeat principle. Robbery cases in six districts of Beijing inner city from 2012 to 2014 were used to examine the extent to which repeats and near repeats spatially intersect robbery hotspots. All the case chains within crime hotspots satisfying repeat and near repeat principle were screened out. From this, by dividing the case characteristics into criminal factors and environmental factors, the characteristics of case chains were analyzed to describe the features and formation of hotspots. The results suggest that there were three main crime hotspots in the six districts of Beijing inner city, namely "a", "b", "c", and most of the cases located within the hotspots were repeats and near repeats. The hotspot "a" was located in Shuangjing and Jinsong, and the hotspot "c" was located in Dahongmen Bridge. The characteristics of the criminal factors of these two hotspots were more consistent than that of the environmental factors, which indicates that the formation of the hotspots were more likely to originate from the repeated crimes committed in the area by criminals. The hotspot "b" was located in Fenzhongsi area by the southeastern third ring road. The characteristics of the environmental factors of this hotspot were more consistent than that of the criminal factors, which indicates that the formation of the hotspot was more likely to originate from different criminals committing crimes in this area. The research findings presented in this article can aid decision making on crime prevention and detection in policing.
Key words: near repeat principle; crime hotspot; Simpson index; case chain; Beijing
表1 时间重复及时空临近重复案件点对数量Tab.1 The number of case pairs |
空间间隔/m | 时间间隔/d | |||||||
---|---|---|---|---|---|---|---|---|
(0, 7] | (8, 14] | (15, 21] | (22, 28] | (29, 35] | (36, 42] | (43, 49] | >49 | |
同一地点 (1, 1000] (1001, 2000] (2001, 3000] (3001, 4000] (4001, 5000] >5000 | 182 59 42 70 106 98 10723 | 46 23 39 89 81 121 11182 | 43 32 42 50 63 87 11291 | 33 11 45 59 73 95 11114 | 42 19 33 58 98 104 11011 | 32 19 53 62 79 90 11184 | 36 33 59 59 69 78 10852 | 2351 989 2691 3918 5349 6260 772976 |
表2 实际案件分布的临近重复计算值与模拟平均期望值之比及其显著性水平Tab.2 Observed over mean expected frequencies and significance levels |
空间间隔/m | 时间间隔/d | |||||||
---|---|---|---|---|---|---|---|---|
(0, 7] | (8, 14] | (15, 21] | (22, 28] | (29, 35] | (36, 42] | (43, 49] | >49 | |
同一地点 (1, 1000] (1001, 2000] (2001, 3000] (3001, 4000] (4001, 5000] >5000 | 5.11** 3.87** 1.09 1.24* 1.40** 1.10 0.98 | 1.25 1.47* 0.98 1.54* 1.03 1.32** 0.99 | 1.17 2.00** 1.06 0.86 0.80 0.95 1.00 | 0.92 0.70 1.14 1.03 0.94 1.04 1.00 | 1.17 1.23 0.84 1.02 1.27* 1.15 1.00 | 0.87 1.21 1.33* 1.08 1.01 0.99 1.00 | 1.02 2.19** 1.54* 1.05 0.91 0.88 1.00 | 0.94 0.92 0.99 0.99 0.99 0.99 1.00* |
注:*、**分别表示通过显著性水平为0.05、0.001的检验。 |
表3 时间重复、时空临近重复案件链数量与占比情况分布Tab.3 The number and proportion of case chains |
统计项 | 链中案件数量 | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | |
时间重复案件链数量/条 时空临近重复案件链数量/条 案件链数量总计/条 在所有案件链中占比/% | 62 52 114 60.6 | 8 14 22 11.7 | 6 12 18 9.6 | 1 16 17 9.0 | 0 16 16 8.5 | 0 0 0 0 | 0 0 0 0 | 0 0 0 0 | 0 0 0 0 | 1 0 1 0.5 |
表4 犯罪热点内案件、案件链数量对应Tab.4 The number of cases and case chains located within crime hotspots |
案件/案件链 | 热点“a” | 热点“b” | 热点“c” |
---|---|---|---|
时间重复案件 时间重复案件链 时空临近重复案件 时空临近重复案件链 | 9 4 17 14 | 9 2 23 50 | 6 2 8 4 |
表5 犯罪人及环境因素案件特征分类Tab.5 Categories of case characteristics of criminal and environmental factors |
案件因素类型 | 案件特征类型 | 案件特征子类 |
---|---|---|
犯罪人因素 | 发案时间 | 凌晨/早晨/上午/中午/下午/晚上 |
作案手段 | 踢打/言语胁迫/扼颈/持锐器/仗势胁迫/露械胁迫/强拿硬要/持钝器/搜身 | |
环境因素 | 发案处所 | 宾馆/高层楼房/商业区/居民小区/网吧/普通楼房/楼群/贸易市场/一般公路/乡村公路/广场/结合部/出租屋/树林/商店/建筑工地/平房/街巷 |
发案部位 | 客房/电梯间/路旁/楼道/胡同/卧室/河边/地下室/街心公园/门前/出租房屋/工地料场/楼口/路口/车厢 | |
被抢物品类型 | 卡/电子产品/钱/包/首饰/日用品/证件/车/钥匙/衣物 |
表6 热点各项案件特征的Simpson指数Tab.6 Simpson index of case characteristics of crime hotspots |
热点编号 | 统计指标 | 犯罪人因素 | 环境因素 | ||||
---|---|---|---|---|---|---|---|
发案时间 | 作案手段 | 发案处所 | 发案部位 | 被抢物品类型 | |||
“a” | 最大值 最小值 平均值 标准差 | 1.000 0 0.648 0.431 | 1.000 0 0.444 0.435 | 1.000 0 0.074 0.244 | 0.167 0 0.019 0.054 | 1.000 0 0.148 0.317 | |
“b” | 最大值 最小值 平均值 标准差 | 1.000 0 0.287 0.196 | 0.600 0 0.146 0.136 | 0.600 0 0.085 0.089 | 1.000 0 0.347 0.192 | 1.000 0 0.447 0.269 | |
“c” | 最大值 最小值 平均值 标准差 | 1.000 0 0.800 0.443 | 1.000 0 0.333 0.365 | 0.333 0 0.067 0.139 | 0 0 0 0 | 1.000 0 0.200 0.400 |
[1] |
|
[2] |
徐嘉祥, 陈鹏, 陈建国 . 基于环境犯罪学理论的入室盗窃时空分布研究: 以北京市主城区案件的分析为例 [J]. 人文地理, 2018,33(1):43-50.
[
|
[3] |
|
[4] |
|
[5] |
冯健, 黄琳珊, 董颖 , 等. 城市犯罪时空特征与机制: 以北京城八区财产类犯罪为例 [J]. 地理学报, 2012,67(12):1645-1656.
[
|
[6] |
朱艳丽, 靖常峰, 伏家云 , 等. 时空立方体的抢劫案件时空特征挖掘与分析 [J]. 测绘科学, 2019,44(9):132-138, 145.
[
|
[7] |
|
[8] |
|
[9] |
毛媛媛, 戴慎志 . 犯罪空间分布与环境特征: 以上海市为例 [J]. 城市规划学刊, 2006(3):85-93.
[
|
[10] |
柳林, 张春霞, 冯嘉欣 , 等. ZG市诈骗犯罪的时空分布与影响因素 [J]. 地理学报, 2017,72(2):315-328.
[
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
[20] |
|
[21] |
|
[22] |
|
[23] |
|
[24] |
|
[25] |
徐冲, 柳林, 周素红 . DP半岛街头抢劫案件的临近重复发生模式 [J]. 地理研究, 2015,34(2):384-394.
[
|
[26] |
Tompson L. Street robbery [EB/OL]. JDiBrief Series, 2012 [ 2019- 10- 02]. http://www.ucl.ac.uk/jdibrief/crime/street-robbery.
|
[27] |
陈鹏, 李欣, 胡啸峰 , 等. 北京市长安街沿线的扒窃案件高发区分析及防控对策 [J]. 地理科学进展, 2015,34(10):1250-1258.
[
|
[28] |
李存华, 孙志挥, 陈耿 , 等. 核密度估计及其在聚类算法构造中的应用 [J]. 计算机研究与发展, 2004,41(10):1712-1719.
[
|
[29] |
禹文豪, 艾廷华 . 核密度估计法支持下的网络空间POI点可视化与分析 [J]. 测绘学报, 2015,44(1):82-90.
[
|
[30] |
|
[31] |
|
[32] |
|
[33] |
|
[34] |
|
[35] |
韩宁, 陈巍 . 基于聚类分析的串并案研究 [J]. 中国人民公安大学学报(自然科学版), 2012,18(1):53-58.
[
|
[36] |
|
/
〈 |
|
〉 |