Constructing Probabilistic Human Mobility Patterns for Estimating Total Exposures and Pathogen Transmission Using...

Area of Science:

• Epidemiology
• Computational Science
• Data Science

Background:

• Aggregated traffic data, like that from Advan, is underutilized for public health research.
• Raw phone location data raises ethical and legal concerns for research and surveillance.
• Existing mobility models may not capture realistic human movement patterns at desired resolutions.

Purpose of the Study:

• To develop a stochastic algorithm for modeling probabilistic human mobility patterns using aggregated traffic data.
• To create dynamic mobility models applicable to exposure and epidemic spread simulations.
• To demonstrate the integration of these models with external exposure fields and epidemiological models.

Main Methods:

• Utilized aggregated traffic data from Advan to develop a stochastic algorithm.
• Modeled human mobility by associating residences and workplaces with unique census blocks.
• Developed a probabilistic exposure model and integrated it with a SIR contagion model.

Main Results:

• The algorithm generates realistic human mobility patterns at weekly or monthly resolutions.
• The developed models are dynamic and can be coupled with various exposure and transmission mechanisms.
• Demonstrated the application in probabilistic exposure modeling and SIR contagion simulations.

Conclusions:

• Aggregated traffic data can be effectively used to model human mobility for public health research.
• The proposed stochastic algorithm provides a flexible framework for epidemiological modeling.
• This approach offers a privacy-preserving alternative to raw phone location data for mobility pattern analysis.