Exploring the mobility of mobile phone users

Abstract

Mobile phone datasets allow for the analysis of human behavior on an unprecedented scale. The social network, temporal dynamics and mobile behavior of mobile phone users have often been analyzed independently from each other using mobile phone datasets. In this article, we explore the connections between various features of human behavior extracted from a large mobile phone dataset. Our observations are based on the analysis of communication data of 100000 anonymized and randomly chosen individuals in a dataset of communications in Portugal. We show that clustering and principal component analysis allow for a significant dimension reduction with limited loss of information. The most important features are related to geographical location. In particular, we observe that most people spend most of their time at only a few locations. With the help of clustering methods, we then robustly identify home and office locations and compare the results with official census data. Finally, we analyze the geographic spread of users' frequent locations and show that commuting distances can be reasonably well explained by a gravity model. Keywords:Human Mobility, Data Mining, Location Detection, Commuting Distance AbstractMobile phone datasets allow for the analysis of human behavior on an unprecedented scale. The social network, temporal dynamics and mobile behavior of mobile phone users have often been analyzed independently from each other using mobile phone datasets. In this article, we explore the connections between various features of human behavior extracted from a large mobile phone dataset. We show that clustering and principal component analysis allows for a significant dimension reduction with limited loss of information. The most important features are related to geographical location. In particular, we observe that most people spend most of their time at only a few locations. of users' frequent locations and show that commuting distances can be reasonably well explained by a gravity model.

On the properties of human mobility

“…To solve the aforementioned issue, Isaacman et al [19] have proposed a technique based on clustering and regression to identify important places then assign them a semantic such as home and work. By contrast, Csaji et al [12] have combined principal component analysis with clustering to robustly identify home and work places. Finally, Arai and Shibasaki [6] have proposed a methodology for the estimation of home and work locations based on time windows.…”

“…In Figure 1 we report a small sample for each kind of recorded activity accompanied by a mobility trace that comes from combining the CDR entries. One of the advantages of this dataset with respect to other datasets [17,10,19,3,12] is the chance to leverage the Internet access data for purposes of mobility pattern analysis [4]. Although CDRs are rich sources for studying and analyzing human activities in different fields, they have two significant drawbacks as to providing location information.…”

Data from mobile phone operators: A tool for smarter cities?

“…The meaning of important activity places is analyzed from different statistical approaches and data sets: Paper based travel behaviour data (Flamm and Kaufmann, 2006); WiFi (Kang et al, 2004); GPS (Zhou et al, 2007); GSM in combination with GPS (Nurmi and Koolwaaij, 2006;Nurmi and Bhattacharya, 2008), and CDR's (Ahas, 2006b(Ahas, , 2008bGonzáles et al, 2008;Bayir et al, 2010;Isaacman et al, 2011a;Csáji et al, 2012;Ranjan et al, 2012). Place learning algorithms can be divided into two classes: geometry and fingerprint.…”

“…For an extensive review on different methodological perspectives, we refer to Nurmi and Bhattacharya (2008) and Ahas et al (2010). Csáji et al (2012) concluded that most people spend most of their time at a few locations 8 . Huang et al (2010) stated that these anchor points 9 and the routes between them are of significant value to effective network management, public transportation planning and city management.…”

“…Since 2006, a number of studies has been initiated to analyze human mobility patterns (e.g. Eagle and Pentland, 2006;Mateos and Fisher, 2006;Shoval, 2007;Gonzalez et al, 2008;Park et al, 2010;Song et al, 2010aSong et al, , 2010bCsáji et al, 2012;Hoteit et al, 2014). For example, Gonzalez et al (2008) conclude that human trajectories show a high degree of temporal and spatial regularity, and follow simple reproducible patterns.…”