基于FP-Growth算法的运毒嫌疑车辆智能推荐研究  

Intelligent Recommendation Regarding Vehicles Suspected of Drug Transportation Based on FP-Growth Algorithm

作　　者：陈柏翰 罗安飞 Chen Bohan;Luo Anfei(Guizhou Police College,Guiyang 550005,China)

机构地区：[1]贵州警察学院,贵州贵阳550005

出　　处：《贵州警察学院学报》2024年第3期84-91,共8页Journal of Guizhou Police College

基　　金：贵州省教育厅青年科技人才成长项目:黔教合KY字[2021]282;融合线上线下项目驱动式应用型教学研究(2022XJJG07)。

摘　　要：毒品运输是毒品犯罪的重要环节,虽然毒品运输的手段越来越多样化,但公路运输仍然是主要的运输方式之一,而运毒人员有着各自经典的运毒模式。文中对运毒模式进行特征挖掘,发现存在前后车伴随的规律,根据实际业务中前后车行为以半小时为时间间隔导向,建模时选择PostgreSQL数据库。在数据库中建立过往车辆前半小时中间表、后半小时中间表、中间跨度表,运用人工智能数据挖掘技术实现从大量的通行车辆中抽取车辆伴随信息,采用FP-Growth算法挖掘频繁项集,查找高频出现车牌号,通过设定阈值并找到对应的关联规则,经过缉毒民警提供的黑名单进行过滤并排序,最后进行车辆嫌疑度的推荐,为民警拦截嫌疑车辆提供支持,能够在一定程度上提高对嫌疑车辆排查的针对性、准确性和有效性。The transportation of narcotic drugs serves as a crucial link in drug-related crimes.Despite the increasing diversity in transportation methods,highway remains the primary route for trafficking narcotic drugs,and different drug traffickers tend to employ distinctive patterns of drug transportation.This study conducts feature mining on drug transportation patterns,unveiling a correlation between the leading vehicle and the following vehicle.After analyzing and summarizing the characteristics of drug transportation on highways,it has been observed that in practice,there is a typical half-an-hour time interval between a leading vehicle and a following vehicle.Based on this finding,the modeling utilizes a PostgreSQL database,wherein three tables are designed for passing vehicles:the first half-hour intermediate table,the second half-hour intermediate table,and the intermediate span table.The technique of artificial intelligence data mining is applied to extracting the information on vehicle following patterns from a large volume of passing vehicles.Subsequently,the FP-Growth algorithm is utilized for mining frequent item sets with the objective of identifying frequently occurring license plates.By establishing thresholds and finding association rules,as well as filtering and sorting blacklists provided by anti-drug law enforcement officers,we can potentially provide recommendations regarding the suspicion level of drug transportation vehicles.This will support police's operation in intercepting vehicles suspected of drug trafficking,thereby enhancing the pertinence,accuracy,and effectiveness of suspect vehicles screenings to a certain extent.

关 键 词：毒品运输 运毒模式 特征挖掘 FP-GROWTH算法 关联规则 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论] D918.2[自动化与计算机技术—计算机科学与技术]