Times are displayed in (UTC-05:00) Eastern Time (US & Canada)Change
Understanding demographic data bias in mobile location data
Topics:
Keywords: human mobility, mobile location data, data bias, mitigation Abstract Type: Paper Abstract
Authors:
Jessica Embury, San Diego State University, UC Santa Barbara
,
,
,
,
,
,
,
,
,
Abstract
Mobile location data is collected by third-party applications on smartphones and other location-enabled devices, and is typically distributed through private companies (e.g., SafeGraph, Spectus). The use of mobile location data has become commonplace for diverse applications. However, the inherent demographic biases present in the data remain understudied and rarely mitigated. While existing literature on the subject recognizes the underrepresentation of lower income and lower education individuals, many of the finer nuances are not well understood. This study extended existing research on the topic of demographic bias in mobile location data through a multi-scale comparison of demographic values from the U.S. Census Bureau’s American Community Survey with values calculated using mobile device home location data from two data sources. The study also investigated whether the combination of the two mobile location datasets can effectively mitigate the detected biases.
Understanding demographic data bias in mobile location data