基于关联规则的道路交通事故多发点成因分析  

Analysis on the Causes of Traffic Accident Prone Points Based on Association Rules

作　　者：王颖志[1] 毛永波[1] 王立君 WANG Yingzhi;MAO Yongbo;WANG Lijun(Department of Traffic Management Engineering,Zhejiang Police College,Hangzhou 310053,China;Geographic Information Science Institute,Zhejiang University,Hangzhou 310028,China)

机构地区：[1]浙江警察学院交通管理工程系,浙江杭州310053 [2]浙江大学地理信息科学研究所,浙江杭州310028

出　　处：《中国人民公安大学学报（自然科学版）》2021年第1期56-62,共7页Journal of People’s Public Security University of China(Science and Technology)

摘　　要：分析道路交通事故成因有助于改善道路交通环境、减少道路交通事故的发生。近年来,关联规则模型及其扩展在道路交通事故多发点成因分析中使用广泛。然而,使用关联规则中传统的兴趣度度量方法和Apriori算法存在许多局限。在传统兴趣度度量方法和Apriori算法的基础上做出改进,使用频数分析法和改进的关联规则方法对事故多发点进行成因分析,并通过对方法实现的难易程度、分析效率和结果的有效性三方面进行对比,以使挖掘结果尽可能地贡献于事故多发点的成因分析。Analyzing the causes of road traffic accidents is helpful to improve the traffic environment and reduce the traffic accidents.In recent years,the Apriori algorithm and its extended model are widely used in the scene of cause analysis of accident prone points.However,there are many limitations when use traditional interest measurement methods and Apriori algorithm.this paper improve the traditional interest measurement method and Apriori algorithm using frequency analysis method and improved association rule method to analyze the causes of accident prone points.Evaluating three aspects of the method,difficulty,analysis efficiency and effectiveness,to make the mining results more contribute to the cause analysis of accident prone points.

关 键 词：道路交通事故多发点 关联规则 成因分析 频数分析 APRIORI算法 

分 类 号：D631.5[政治法律—政治学]