基于聚类关联规则的公交扒窃犯罪时空分析

发布时间：2018-03-06 02:34

  本文选题：公交扒窃　切入点：聚类分析　出处：《华东师范大学学报(自然科学版)》2017年03期 　论文类型：期刊论文

【摘要】：提出了一种基于聚类的时空关联规则的公交犯罪挖掘算法.针对某市一个区的110报警数据库中的大量业务信息进行分析.首先,通过文本挖掘技术从案情信息中提取时间、地点等信息,并利用高德地图API的地理编码服务和POI搜索功能对提取的地址信息进行地址匹配,提取受害人上下车站点、乘坐公交线路等信息.其次,对提取得到的时空数据进行归并处理.最后,根据案发时段、季节以及是否节假日进行聚类分析,然后在簇内进行时空关联规则分析.这种挖掘方法具有以下特点:①在聚类基础上进行关联规则分析,减少扫描数据库次数,大大缩小数据扫描范围,提高算法效率,更加适合海量犯罪数据的挖掘.②聚类后簇内数据具有相似性,特征更加明显,在此基础上进行关联规则分析产生较小的频繁项集,并且提取出置信度较高的规则.③考虑犯罪行为的时空特性,挖掘过程中同时考虑了案发季节、是否节假日等因素.
[Abstract]:This paper presents an algorithm of public transportation crime mining based on clustering space-time association rules. It analyzes a lot of business information in 110 alarm database of a certain city. Firstly, time is extracted from the case information by text mining technology. Location and other information, and using Amap API's geo-coding service and POI search function to match the extracted address information, extract information such as the victim's boarding and alighting station, bus route, etc. Secondly, Finally, clustering analysis is carried out according to the time of the crime, the season and whether the holiday or not. Then the time-space association rule analysis is carried out in the cluster. This mining method has the following characteristics: 1: 1 analyzes association rules on the basis of clustering, reduces the number of scanning databases, greatly reduces the scope of data scanning, and improves the efficiency of the algorithm. It is more suitable for mining large amount of crime data. 2. After clustering, the data in cluster have similarity and more obvious features. On this basis, association rule analysis produces smaller frequent itemsets. The rule of high confidence is extracted to consider the temporal and spatial characteristics of criminal behavior, and the factors such as the crime season, whether the holiday or not are considered in the mining process.
【作者单位】： 华东师范大学地理科学学院;
【基金】：国家理科基地科研训练及科研能力提高项目(J1310028)
【分类号】：D917.3;TP311.13
，

本文编号：1572981