Imputing Socio-Demographics for Mobile Trajectors

Term Start:

June 1, 2024

Term End:

May 31, 2025

Budget:

$332,045

Keywords:

Big Data, Mobile Devices, Mobility Patterns, Passive Data

Thrust Area(s):

Data Collection Mechanisms, Equity and Understanding User Needs

University Lead:

University of Washington

Researcher(s):

Cynthia Chen

Ubiquitous mobile devices have resulted in massive amount of location- and time-stamped traces that can be used to infer people’s mobility patterns for various applications. Unlike household travel survey data that is small but rich (short but wide data), mobile data, is often massive but shallow (long but thin data) whose meanings in terms of people’s travel patterns must be inferred. Not only being massive, it is also longitudinal, or to be precise: the data is continuous. These two key features hold great promises for a wide range of applications that cannot achieved with the traditional household travel survey data. Examples include: just in time or real time policy evaluations, a closed-loop from real time demand forecasting to service provision and then back to demand monitoring, and creation of digital twins for whole-city simulations. This study addresses a critical challenge that needs to be overcome in order to realize the great promises that the big, passively-generated mobile data offers. That is: to impute socio-demographics from the census data with the mobile trajectories generated from the big data. The novelty of the proposed project lies in that the proposed model will explicitly recognize the uncertainty that exists in the linkage between socio-demographics and travel behaviors.

Scroll to Top