关键词: 手机信令数据, 时空分布, 移动模式

Abstract: Aiming at the problems of data sparseness, uneven distribution and abnormal signal drift in the mobility pattern studies based on mobile phone signaling data, this paper proposes a data preprocessing method to solve data sparseness and uneven distribution from two levels：data capacity and data distribution. A spatial and temporal distribution algorithm of the people flow based on main active area is used to solve the shortcomings of the existing researches in considering the difference of time slots. Taking Shanghai as an example, the spatial and temporal distribution of the people flow in different functional areas on weekdays and weekends is compared and analyzed. Finally, a stay point extraction algorithm based on grid which considers both space and time factors is proposed to study the daily mobility patterns of humans based on residence. The experimental results show that the algorithm can extract meaningful stay points for users more accurately and obtain more concise user mobility patterns.

Key words: mobile phone signaling data, spatial and temporal distribution, mobility patterns