城市居民时空行为序列模式挖掘方法

Data Mining Method for Exploring Spatio-Temporal Behavior Sequence Patterns of Urban Residents

通过问卷调查获得北京市500个家庭的活动日志数据。运用序列比对方法对时空行为序列数据进行序列模式挖掘,对每类序列模式通过频繁模式挖掘出其中潜在有用的行为模式。试验结果表明,序列比对方法与频繁模式挖掘相结合在城市居民行为模式分析中应用成功,其弥补了传统数据分析方法的不足,为复杂时空行为数据的分析提供了一种新视角。

This paper proposed to introduce sequence alignment and data mining method to analyze temporal and spatial behavior patterns of urban residents.The experimental data are activity log data of 500 families through questionnaires about daily activity in Beijing.The sequence alignment method was firstly used to find groups with similar behavior patterns.Then frequent pattern mining method was used to find meaning associated attributes for each behavior group.The experimental results showed that sequence alignment combined with frequency pattern mining method provides a new perspective in spatio-temporal behavior data analysis.