Assessing and Mitigating Spatial Bias of Large-Scale Mobile Location Data for Human Mobility Analysis

University Of Florida, Gainesville FL

Investigators

Abstract

The availability of large-scale mobile location data, which can track millions of people’s movements over time with high spatial resolutions, is transforming human mobility research. However, the potential spatial biases contained in such data, which reflect systematic errors or inaccuracies in geographic information, pose a significant challenge to the reliability and validity of research findings derived from them. This project tackles this challenge by advancing knowledge on the causes and extent of spatial biases associated with large-scale mobile location data and developing methods to mitigate these biases. It contributes to theoretical and methodological advances in geographical and behavioral sciences and their intersection in the context of human mobility research. The research findings advance STEM education and inform transportation agencies that use mobile location data to develop more equitable transportation plans and policies. This project supports research on quantifying, identifying the causes of, and mitigating potential spatial biases in large-scale mobile location data used for human mobility analysis. It addresses four major research tasks: 1) Developing metrics and analytical frameworks to evaluate bias, collecting multi-sourced datasets, and then quantifying spatial bias in sample representation and mobility measurements; 2) conducting mixed-methods research to identify potential causes of spatial biases from the mobile location data generation process and assessing how algorithmic uncertainties can result in spatial bias; 3) developing new methods to mitigate spatial bias in mobile location data; and 4) publishing research products, building partnerships, and promoting results adoption. The methods produced can be generalized to address spatial biases in other emerging datasets such as geo-tagged social media data and participatory Geographic Information Systems data. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.