Journey reconstruction based on social media posts. A study on data quality and privacy implications

Abstract

Studies of journeys can provide insights in the way in which movement influences the world socially and physically. However, journey data is difficult to obtain, as it entails private events over a longer period of time. Fortunately, smartphones with GPS functionality and social media usage have increasingly become a common part of daily life, for which reason geotagged personal posts from social media can serve as input journey data. Therefore, the objectives of this research are to assess whether or not social media personal posts are of sufficient quality for journey reconstruction, and to provide insight in the relation between quality of journey reconstruction and the associated privacy risk. This quality of personal posts for journey reconstruction is assessed by obtaining social media post histories from participants' Instagram accounts. From these post histories geotags and timestamps are extracted, that serve as input for individual journey reconstruction. This journey reconstruction is personally evaluated per participant based on six quality dimensions. The evaluation of journey reconstructions reveals that the better the quality of the journey reconstruction, the higher the risk of location privacy. Furthermore, temporal attributes are most influential on the quality of the journey reconstruction, because inconsistencies in the temporal attributes disarrange the sequence of stop places, that complicates the process of journey reconstruction.