Building Infectious Disease through Prediction Algorithms and Data Ownership

Authors: Anna Beck*, University of Wisconsin
Topics: Cultural and Political Ecology, Economic Geography, Cultural Geography
Keywords: infectious disease, algorithms, property, biosecurity, data, health
Session Type: Paper
Day: 4/4/2019
Start / End Time: 1:10 PM / 2:50 PM
Room: Palladian, Omni, West
Presentation File: No File Uploaded


In 2008 Google announced Google Flu Trends (GFT) with the goal of using web search queries to track and predict ILI (influenza like illness) and emerging infectious diseases like Dengue and Ebola. My research focuses on how the goal of prediction shapes what parameters are important in ILI detection. Who owns the data needed to provide real time or predictive insights into ILIs also plays a large role in what becomes important when detecting ILIs. This paper will explore how Google’s ownership of search engine query data and population movement data has influenced what is important to modelers and algorithm builders who develop infectious disease prediction technologies. This is true both for Google’s collaborators who have access to their data and those who don’t. In infectious disease prediction literatures search engine query data coupled with location is the gold standard for a functional predictive model, gesturing at a cemented digital enclosure. However, reactions in Europe have led to crowd sourced methods of tracking and predicting ILI outbreaks (e.g. Flu Near You). Groups that have access to open and closed source cellular data have also leveraged movement and mobility data to build competing algorithms while arguing that search engine query data has significant flaws. In each case who owns and controls information pertaining to the spread and emergence of ILIs effects what is made important about the ways ILIs develop and move through populations.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login