Uncovering the relationship between point-of-interests-related human mobility and socioeconomic status. (June 2019)

• Record Type: Journal Article
• Authors: Li, Dong; Liu, Jiming
• Is Part Of: Telematics and informatics. Volume 39(2019)
• Page Start: 49
• Page End: 63
• Publication Date: 2019-06
• Subjects: Human mobility -- Urban socioeconomic -- Machine learning -- Prediction -- Big data; Telecommunication -- Periodicals; Computer networks -- Periodicals; Télécommunications -- Périodiques; Réseaux d'ordinateurs -- Périodiques; 384
• DOI: 10.1016/j.tele.2019.01.001
• Languages: English
• ISSNs: 0736-5853

Abstract

Highlights: There exist obvious correlations between the POIs-related human mobility and the socioeconomic indicators at city level. Different socioeconomic indicators have intrinsic relevance from the perspective of POIs-related human mobility. A multi-task prediction framework based on POIs-related human mobility is proposed for forecasting socioeconomic indicators. The developed framework can predict socioeconomic indicators with high accuracy, and perform much better than the traditional signal-task method. Abstract: In a city or region, understanding the relationship between human mobility and socioeconomic status is critical to public policies formulation, urban design and marketing strategies development. Based on the available massive geo-located human data, existing studies focused almost exclusively on the position attributes ( i.e. coordinates) of the locations visited by people to explore the relationship, however, they ignored the category attributes ( e.g. restaurant or supermarket) of these locations which imply the purposes ( e.g. eating or shopping) behind human movements. A location with coordinates and category information is usually referred to as a point-of-interest (POI). In this paper, we study the relationship between POIs-related human mobility and socioeconomic status at city level. Starting from the location-based social network ( i.e. Foursquare) dataset, we find that the check-in numbers of location categories are correlated with socioeconomic indicators, either positively or negatively. To further validate these correlations, we develop and test a multi-task prediction framework based on POIs-related human mobility for forecasting socioeconomic indicators. Extensive experiments on the Foursquare dataset show that the socioeconomic indicators can be well predicted by our proposed framework. Our findings and methods are helpful for modeling human mobility and assessing socioeconomic status.