摘 要:群体事件预测对群体事件管理具有重要作用。文章通过分析时空轨迹数据的特点,首先确定使用频繁模式对多维度时空轨迹进行数据挖掘,其次对时空轨迹数据进行预处理,最后通过优化Apriori 算法,提出一种MapReduce 框架下基于Apriori 算法的时空轨迹数据挖掘算法,实现在并行运算环境下对时空轨迹数据进行关联规则挖掘。实验表明:该算法可以找出经济群体事件发生的关键因素特征,这些特征值符合不同类型经济事件的特点,为公安行业经济群体事件管理提供决策部署依据。
关键词:群体事件;时空轨迹;Apriori 算法;并行运算;数据挖掘
中图分类号:TP311.13 文献标识码:A 文章编号:2096-4706(2019)01-0078-03
Application of Spatio-Temporal Trajectory Data Mining in the Police Management of Economic Group Events
WANG Zhongni1,JIN Tao2
(1. Public Security Department of Shanxi Province,Taiyuan 030006,China;2.Taiyuan Fire Brigade,Taiyuan 030006,China)
Abstract:Group event prediction plays an important role in the management of group events. This paper analyzes the characteristicsof spatio-temporal trajectory data. Firstly,by comparison,the frequent patterns are the optimal methods for the multi-dimensional spatiotemporaltrajectory data mining. Secondly,the trajectory data was pre-processed. Finally,this paper proposes a new spatio-temporaltrajectory data mining algorithm based on MapReduce framework. The experiment proves that the new algorithm can find out the keyfactors which affect the occurrence of economic group events. And the factors are consistent with the characteristics of different types ofeconomic events. It can provide decision-making basis for the management of economic group events.
Keywords:group event;spatial-temporal trajectory;Apriori algorithm;parallel computing;data mining
参考文献:
[1] 陈鹏,赵鹏凯,瞿珂. 基于时间序列模型的110 警情数据预测研究 [J]. 信息系统工程,2015(9):128-131.
[2] 张圣. 一种基于云计算的关联规则Apriori 算法 [J]. 通信技术,2011,44(6):141-143.
[3] 邱昕,甘超,江雄心,等. 基于云计算环境下Apriori 算法的设备故障诊断技术研究 [J]. 组合机床与自动化加工技术,2014(4):45-48.
[4] 吉根林,赵斌. 时空轨迹大数据模式挖掘研究进展 [J]. 数据采集与处理,2015,30(1):47-58.
[5] 彭曲,丁治明,郭黎敏. 基于马尔可夫链的轨迹预测 [J].计算机科学,2010,37(8):189-193.
作者简介:王仲妮(1983.12-),女,汉族,山西宁武人, 副主任科员,硕士,研究方向:大数据挖掘。