Mobility network models of COVID-19 explain inequities and inform reopening

This seminar was led by Serina Chang, who is currently pursuing a PhD in Computer Science (CS) at Stanford University. Her research focuses on developing methods to model and make predictions over complex systems of human behavior. To do this, she leverages experience in epidemic modeling, (dis)information dynamics, political polarization, and recommendation systems.

Understanding contact is crucial to understanding how the COVID-19 virus spreads, but how can we estimate this in a reliable way with so many variables?

Chang’s recent research project seeks to answer this (and more!) by examining the underlying network of people and places that modulate the spread of the virus. The seminar summarizes their findings from their paper: Mobility network models of COVID-19 explain inequities and inform reopening.

You can also learn more about the work they are doing on COVID-19 Mobility Modeling!

Modeling Mobility Networks of COVID-19 to Identify Inequities and Plan Reopening

Using aggregate data from cell phone devices provided by SafeGraph, researchers can reliably capture spread based on contact patterns. Using this data, Chang and her peers were able to develop a model of mobility patterns that allowed them to understand where and when individuals got infected, identify the neighborhoods at the highest risk, and determine which POIs are the riskiest places for interaction.

This information also allows for future planning, as they can modify any input in their model, accounting for different factors and variables. With these models, they can test minute changes to mobility patterns and test vaccination strategies for the best outcomes. This allows them to then predict and model inequities and reopenings with accuracy, reliability, and confidence.

How Placekey enabled research on COVID-19 reopenings and inequities from a mobility network perspective

SafeGraph’s anonymized, aggregate location data from numerous cell phone apps enabled this research. With detailed information about points of interest (POIs), including the hourly number of visits, median dwell time, and category, they can paint a vivid picture of mobility patterns and how different CBGs interact with the places around them.

Using census block group (CBG) data, they can analyze the demographic makeup of these groups, and ultimately examine how different groups interact with each other. By combining POI and CBG information, they can create a network that allows them to model this mobility information, helping to visualize and assess findings.

To learn how your team can benefit from using SafeGraph and Placekey, join the SafeGraph community. There is an entire community of users passionate about the work they are doing that can share their own strategies and solutions for how to use Placekey and SafeGraph.