Big data-enabled social sensing in spatial analysis: Potentials and pitfalls

Over the past decade, there has been an unprecedented boom in the application of big data derived from location-aware technology in GIScience. However, reflections on data quality lead to concerns about inaccurate or even erroneous spatial analysis results. This article contributes a meta-analysis of big data-enabled social sensing in spatial analysis. Big data-enabled research has substantially enhanced our understanding of human population distribution and dynamics, land use, travel behavior, movement patterns, and spatial social networks from city to country to global levels. However, the effectiveness of big data-enabled social sensing can be undermined by five critical issues related to research design: data volume, spatial coverage, locational information quality, and representativeness across socioeconomic and demographic groups. We suggest a framework for addressing the inherent bias in spatial big data at multiple stages of the research lifecycle and applying them to appropriate topics and scales of applications.