Mobile Homophily and Social Location Prediction

TL;DR

This paper explores how social relationships and personality traits influence human mobility patterns and improves location prediction by leveraging social network data.

Contribution

It investigates the interdependency between mobility and social proximity and demonstrates how social network information can enhance location prediction accuracy.

Findings

Social proximity correlates with similar mobility patterns.

Social network data improves location prediction accuracy.

Personality traits influence mobility predictability.

Abstract

The mobility behavior of human beings is predictable to a varying degree e.g. depending on the traits of their personality such as the trait extraversion - introversion: the mobility of introvert users may be more dominated by routines and habitual movement patterns, resulting in a more predictable mobility behavior on the basis of their own location history while, in contrast, extrovert users get about a lot and are explorative by nature, which may hamper the prediction of their mobility. However, socially more active and extrovert users meet more people and share information, experiences, believes, thoughts etc. with others. which in turn leads to a high interdependency between their mobility and social lives. Using a large LBSN dataset, his paper investigates the interdependency between human mobility and social proximity, the influence of social networks on enhancing location prediction of an individual and the transmission of social trends/influences within social networks.