关键词: 手机信令数据, 个体单元, 高簇聚类, 出行类型, 画像刻画

Abstract: Based on the long-term mobile phone signaling data of Beijing volunteers from Microsoft Research Asia Geolife project from 2007 to 2012, this article conducts research on travel types and portraits based on individuals. This paper proposes a method of character portrait characterization by first classifying user trajectory types based on high-cluster clustering, and then combining travel rules to consider occupation type, age, and hobby attributes. The main conclusions of the study are: ① This paper divides five types of travel including two-point, one-line fixed type, two-point, one-line change type, dual-core type, uniform distribution type, and divergent type. ② Beijing has fixed-work technology people or white-collar workers account for about 44%, and the student group or retired elderly account for nearly a quarter.

Key words: mobile signaling data, individual units, high clustering, travel type, characterization portrait