摘要
[目的/意义]利用数据挖掘技术对恐怖分子的各种日常信息,如购物、社交、交通、通话记录、视频等行为进行分析,对涉恐线索进行预警和排查,越来越成为国际反恐的通用手段之一。如何利用数据挖掘对大量的涉恐基础数据进行快速分类成为当前涉恐情报分析的研究热点。[方法/过程]将研究如何利用基于基尼系数的特征选择方法对涉恐人员的情报信息进行快速分类。分类步骤分为多个层次的属性分裂,其中每个层次中包括三个步骤,分别为计算样本集的基尼系数,计算不同属性的基尼系数,通过比较基尼系数选择分裂属性。[结果/结论]该方法可以对大量涉恐情报基础数据进行快速分类,提高反恐预警的效率。
[ Purpose/Significance ] It becomes one of the common counter-terrorism methods that analyze the various daily lives of terror- ists, such as shopping, socializing, transportation, phone records, video, etc. Data mining technology plays an important role in the field of intelligence analysis. Classification is one of the core application of data mining for counter-terrorism under the environment of big da- ta. [ Method/Process] This paper will propose how to use gini index to quickly classify the information of terrorists. There are many lev- els in the analyzing process. Every level contains three steps -- computing gini index of whole sample set, calculating gini index of each attribute, selecting schismatical attribute. [ Result/Conclusion] This method will increase efficiency of intelligence analysis and improve the mechanism of terror threat warning.
出处
《情报杂志》
CSSCI
北大核心
2017年第4期29-32,53,共5页
Journal of Intelligence
基金
中国人民公安大学基本科研业务费项目"大数据环境下反恐怖情报的数据挖掘分类方法研究"(编号:2015JKF01223)
国家社会科学基金项目"反恐维稳背景下边疆地区维稳战略研究"(编号:14BZZ028)的研究成果之一
关键词
基尼系数
反恐
数据挖掘
决策树
特殊行为轨迹
思想倾向
gini index
counter-terrorism
data mining
decision tree
track of special behavior
ideological inclination
