改进的PrefixSpan算法在客流计数中的应用  

Application of improved PrefixSpan algorithm in counting passenger flow

作　　者：王争[1] 李晋宏[1] 

机构地区：[1]北方工业大学,北京100044

出　　处：《计算机时代》2013年第6期50-52,共3页Computer Era

摘　　要：序列模式挖掘是基于关联规则的频繁项集的挖掘,其实质是在关联模型中加入时间属性。利用改进的PrefixSpan算法对客流计数系统中不同时段的数据进行挖掘分析,给出不同时段的客流高峰的频繁序列模式,对于提高客流计数系统的精度,给管理决策者调配人力,物力,财力提供技术支持,对于最大限度地发掘购物中心的潜力,提高利润,具有重要的经济意义。Sequential pattern mining is a frequent item sets mining based on association rules, and its essence is to add the time attribute to the relation model. In this paper, data in different times of passenger flow counting system is mined and analyzed by the improved PrefixSpan algorithm. Frequent sequential patterns of the peak passenger flow in the different periods are given. It has important economical meaning for improving passenger flow counting system accuracy, providing technical support for the managers to allocate human, material and financial resources, maximizing the potential of the shopping center, and increasing profits.

关 键 词：序列模式挖掘 关联模型 PREFIXSPAN算法 客流计数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]