Abstract: In order to detect stay points from a user’s mobile phone trajectory, this paper makes use of the concept of a functional area to combine the trajectory of mobile phones with geographic information. By considering the user’s overall activities in a particular functional area, this method can identify whether a user stays or not, and avoid misjudgment due to a user’s normal switching and error switching. This method is more general because a stay point is a collection of cells in a continuous trajectory at a functional area instead of a small range covered by a single base station.

Key words: background geographic information, stay point, functional area

摘要： 针对现有停驻点的问题，将手机用户轨迹与背景地理信息相结合，并引入功能区概念，通过考虑用户在某一功能区域内的整体活动来判断用户是否在此停驻。通过对带有标签的模拟的手机轨迹数据进行实验，证明了新的停驻点识别方法有效地避免了由于正常切换、错误切换带来的误判。停驻点的判断更为宽泛，不仅限于一个小的范围（单个基站的覆盖范围），而是由多个相邻且功能相似小区组成的功能区中的小区集合。

关键词: 背景地理信息, 停驻点, 功能区