Data mining techniques can extract useful activity and travel information from large-scale data sources such as mobile phone geolocation data. This paper aims to explore individual activity-travel patterns from samples of mobile phone users using a two-week geolocation data set from the Paris region in France. After filtering the data set, we propose techniques to identify individual stays and activity places. Typical activity places such as the primary anchor place and the secondary place are detected. The daily timeline (i.e., activity-travel program) is reconstructed with the detected activity places and the trips in-between. Based on user-day timelines, a three-stage clustering method is proposed for mobility pattern analysis. In the method framework, activity types are first identified by clustering analysis. In the second stage, daily mobility patterns are obtained after clustering the daily mobility features. Activity-travel topologies are statistically investigated to support the interpretation of daily mobility patterns. In the last stage, we analyze statistically the individual mobility patterns for all samples over 14 days, measured by the number of days for all kinds of daily mobility patterns. All individual samples are divided into several groups where people have similar travel behaviors. A kmeans++ algorithm is applied to obtain the appropriate number of patterns in each stage. Finally, we interpret the individual mobility patterns with statistical descriptions and reveal home-based differences in spatial distribution for the grouped individuals.
Heat exposure assessment based on individual daily mobility patterns in Dhaka, Bangladesh
