Limits of Predictability in Human Mobility

Ci, Song; Qu, Zehui; Blumm, Nicholas; Barabási, Albert László

doi:10.1126/science.1177170

Cited by 2,762 publications

(2,390 citation statements)

References 25 publications

Supporting

Mentioning

2,224

Contrasting

Unclassified

Order By: Relevance

“…Several studies confirmed the intuition that human mobility is highly predictable [9,16], centered around a small number of base locations. This opens a wide range of opportunities for more intelligent recommendations and support of routine activities.…”

Section: Introductionmentioning

confidence: 54%

“…Song et al [16] found that 93% of human mobility is predictable; how predictable an individual's movements is, depends on the entropy of his patterns. However, for predictability it did not make a difference whether an individual's life was constrained to a 10-km neighborhood or whether he travels hundreds of kilometers on a regular basis.…”

Section: Human Mobility Patternsmentioning

confidence: 99%

See 1 more Smart Citation

Predicting User Locations and Trajectories

Herder

Siehndel

Kawase

2014

User Modeling, Adaptation, and Personalization

View full text Add to dashboard Cite

Abstract. Location-based services usually recommend new locations based on the user's current location or a given destination. However, human mobility involves to a large extent routine behavior and visits to already visited locations. In this paper, we show how daily and weekly routines can be modeled with basic prediction techniques. We compare the methods based on their performance, entropy and correlation measures. Further, we discuss how location prediction for everyday activities can be used for personalization techniques, such as timely or delayed recommendations.

show abstract

Section: Introductionmentioning

confidence: 54%

Section: Human Mobility Patternsmentioning

confidence: 99%

Predicting User Locations and Trajectories

Herder

Siehndel

Kawase

2014

User Modeling, Adaptation, and Personalization

View full text Add to dashboard Cite

show abstract

“…This is a justifiable approximation, because of the consensus in the literature is that there is a high predictability in user trajectories within different segments, e.g. [Song et al, 2010], [Lu et al, 2013]. II.…”

Section: It Throws a Warning If Some Antenna Is Overloadedmentioning

confidence: 99%

Fuzzy Recommendations in Marketing Campaigns

Podapati

Lundberg

Sköld

et al. 2017

Communications in Computer and Information Science

View full text Add to dashboard Cite

The population in Sweden is growing rapidly due to immigration. In this light, the issue of infrastructure upgrades to provide telecommunication services is of importance. New antennas can be installed at hot spots of user demand, which will require an investment, and/or the clientele expansion can be carried out in a planned manner to promote the exploitation of the infrastructure in the less loaded geographical zones. In this paper, we explore the second alternative. Informally speaking, the term Infrastructure-Stressing describes a user who stays in the zones of high demand, which are prone to produce service failures, if further loaded. We have studied the Infrastructure-Stressing population in the light of their correlation with geo-demographic segments. This is motivated by the fact that specific geo-demographic segments can be targeted via marketing campaigns. Fuzzy logic is applied to create an interface between big data, numeric methods for processing big data and a manager.

show abstract

“…Census information is less detailed when considered at the individual level, but it has the advantage of covering a significant part of the population of full countries. Recent works analyzing mobile phone records have shown that people spend most of their time in a few locations [22]. These locations are likely to be those registered in the census and, indeed, census-based information has been also used recently to forecast the propagation patterns of infectious diseases such as the latter influenza pandemic [23,24,25].…”

Section: Introductionmentioning

confidence: 99%

“…Human mobility has been studied in recent years with relatively indirect techniques such as tracking bank notes [20] or with more direct methods such as tracking cell phone communications [21,22]. A more classical source of information in this issue is the census.…”

Section: Introductionmentioning

confidence: 99%

Persistence in Voting Behavior: Stronghold Dynamics in Elections

Pérez

Fernández-Gracia

Ramasco

et al. 2015

Social Computing, Behavioral-Cultural Modeling, and Prediction

View full text Add to dashboard Cite

Abstract. Influence among individuals is at the core of collective social phenomena such as the dissemination of ideas, beliefs or behaviors, social learning and the diffusion of innovations. Different mechanisms have been proposed to implement inter-agent influence in social models from the voter model, to majority rules, to the Granoveter model. Here we advance in this direction by confronting the recently introduced Social Influence and Recurrent Mobility (SIRM) model, that reproduces generic features of vote-shares at different geographical levels, with data in the US presidential elections. Our approach incorporates spatial and population diversity as inputs for the opinion dynamics while individuals' mobility provides a proxy for social context, and peer imitation accounts for social influence. The model captures the observed stationary background fluctuations in the vote-shares across counties. We study the so-called political strongholds, i.e., locations where the votes-shares for a party are systematically higher than average. A quantitative definition of a stronghold by means of persistence in time of fluctuations in the voting spatial distribution is introduced, and results from the US Presidential Elections during the period 1980-2012 are analyzed within this framework. We compare electoral results with simulations obtained with the SIRM model finding a good agreement both in terms of the number and the location of strongholds. The strongholds duration is also systematically characterized in the SIRM model. The results compare well with the electoral results data revealing an exponential decay in the persistence of the strongholds with time.

show abstract

Limits of Predictability in Human Mobility

Cited by 2,762 publications

References 25 publications

Predicting User Locations and Trajectories

Predicting User Locations and Trajectories

Fuzzy Recommendations in Marketing Campaigns

Persistence in Voting Behavior: Stronghold Dynamics in Elections

Contact Info

Product

Resources

About