Location uncertainty has been a major barrier in information mining from location data. Although the development of electronic and telecommunication equipment has led to an increased amount and refined resolution of data about individuals’ spatio‐temporal trajectories, the potential of such data, especially in the context of environmental health studies, has not been fully realized due to the lack of methodology that addresses location uncertainties. This article describes a methodological framework for deriving information about people's continuous activities from individual‐collected Global Positioning System (GPS) data, which is vital for a variety of environmental health studies. This framework is composed of two major methods that address critical issues at different stages of GPS data processing: (1) a fuzzy classification method for distinguishing activity patterns; and (2) a scale‐adaptive method for refining activity locations and outdoor/indoor environments. Evaluation of this framework based on smartphone‐collected GPS data indicates that it is robust to location errors and is able to generate useful information about individuals’ life trajectories.

中文翻译：

---

解决基于GPS的活动监视中的位置不确定性：方法框架。

位置不确定性一直是从位置数据挖掘信息的主要障碍。尽管电子和电信设备的发展已导致有关个人时空轨迹的数据的数量增加和分辨率得到了改善，但是由于这种情况，尤其是在环境健康研究的背景下，这些数据的潜力尚未得到充分认识。缺乏解决位置不确定性的方法。本文介绍了一种方法框架，可从个人收集的全球定位系统（GPS）数据中获取有关人们持续活动的信息，这对于各种环境健康研究至关重要。该框架由两种主要方法组成，这些方法解决GPS数据处理不同阶段的关键问题：（1）用于区分活动模式的模糊分类方法；（2）调整活动地点和室外/室内环境的比例缩放方法。根据智能手机收集的GPS数据对该框架进行的评估表明，该框架可防止位置错误，并且能够生成有关个人生活轨迹的有用信息。