地球信息科学学报 ›› 2014, Vol. 16 ›› Issue (4): 537-544.doi: 10.3724/SP.J.1047.2014.00537

• • 上一篇    下一篇

基于加权时空关联规则的公交扒窃犯罪模式识别

叶文菁(), 吴升*()   

  1. 福州大学 福建省空间信息工程研究中心,福州 350001
  • 收稿日期:2014-02-10 修回日期:2014-03-20 出版日期:2014-07-10 发布日期:2014-07-10
  • 作者简介:

    作者简介:叶文菁(1987-),女,福建长泰人,硕士生,研究方向为时空数据挖掘。E-mail:yewenjing06@gmail.com

  • 基金资助:
    国家“863”重大项目课题(2012AA12A208)

Identifying Crime Patterns of Bus Pickpocketing Using Weighted Spatio-Temporal Association Rules Mining

YE Wenjing(), WU Sheng*()   

  1. Spatial Information Research Center of Fujian Province, Fuzhou University, Fuzhou 350001, China
  • Received:2014-02-10 Revised:2014-03-20 Online:2014-07-10 Published:2014-07-10
  • About author:

    *The author: CHEN Nan, E-mail:fjcn99@163.com

摘要:

近年来,公交扒窃案呈上升趋势,为了预防和打击此类犯罪,需要有效识别其犯罪模式。传统的犯罪分析方法,往往将时间和空间分割开来研究,本文则引入加权时空关联规则进行挖掘分析,试图找出公交扒窃的案发时空规律。首先,对公交扒窃数据进行时间粒度和空间粒度的划分,将公交主要运营时间以2 h为单位划分成等间隔的公交时段并对其进行编码,将公交线路按公交站点划分成公交路段;其次,对数据进行空间分析和时间归并,提取出每个案件发生的公交路段和案发时段,并将案发时段归并到公交时段中;再次,由于每个公交路段的案发率不同,其对结果的贡献率也不同,因此,给每个路段赋予一个权重;最后,用Apriori算法进行加权关联规则挖掘,得到公交扒窃的时空犯罪模式。研究表明,这种挖掘方法具有以下特点:(1)按公交站点进行公交路段的划分具有创新性;(2)通过对案发路段的加权,能将空间位置重要程度的差异区分开来,更符合实际情况;(3)挖掘过程中同时考虑了时间与空间属性。

关键词: 时空关联规则, 犯罪模式, 公交扒窃

Abstract:

In recent years, the number of bus pickpocketing is on the rise, which has caused great danger to the public. In order to prevent and combat bus pickpocketing, we need to effectively identify the crime patterns. However, since bus pickpocketing may occur when the bus is moving, it is difficult to define a location and a crime time for it, and the traditional methods of crime analysis often focus on spatial or temporal properties of crimes separately. Therefore, this paper introduced the weighted spatio-temporal association rules mining to find out the spatio-temporal crime patterns of bus pickpocketing. It can be carried out through four steps. Firstly, conduct the temporal and spatial granularity division: divide the main bus running time into equal units of 2h and encode these units, which are entitled by “bus time”. Then, divide the bus routes into sections according to the bus stops, and entitle these sections by “bus section”. Secondly, conduct the spatial analysis and time merge process: extract and gather bus sections and crime time information for pickpocketing cases, and merge the crime time with “bus time”. Thirdly, as the crime rate of each “bus section” varies, their contributions to the result will also be different. Therefore, this paper assigned each “bus section” a weight according to their crime rates, based on the assumption that every “bus section” of a crime case has the same crime rate. Finally, conduct the weighted spatio-temporal association rules mining: use Apriori algorithm to find out the spatio-temporal crime patterns of bus pickpocketing. The results prove that the proposed mining model has the following characteristics: (1) it is innovative to extract bus sections according to the bus stops; (2) It is more realistic to assign a weight to each “bus section” to distinguish the importance of spatial location; (3) The crime patterns show the correspondence between the bus sections and “bus time” of high crime rates. Further work would focus on the efficiency of Apriori algorithm, the evaluation of association rules, the weight settings (consolidating other factors) and the result visualization.

Key words: spatio-temporal association rules, crime pattern, bus pickpocketing