Motivation

This project seeks to define a simple data standard for infectious disease outbreaks, and atop of that provide free software source code which can visualize that information using web technology. This page describes the motivation.

In researching the 2014 Ebola Outbreak in West Africa situation, the main conclusion of this Outbreak Time Series project was that readily available, liberally licensed open source tools for producing visually engaging outbreak dashboards would help mobilize a response earlier and maintain public awareness for the duration of a crisis.

An even deeper problem is the lack machine readable data standards for outbreak data. During the Zika outbreak of 2016, MIT researchers were slowed down by a lack of data standards:

it’s worth noting that these data [from PAHO] aren’t in a machine-readable format (and scraping from interactive visualizations can often be more challenging than doing so from PDFs). This, of course, adds an additional, unnecessary step for researchers and curious minds alike

What millions of people might see is clearly of the greatest impact, but more importantly there is a serious need for a data standard for outbreak data. In other words, there needs to be standard data structures and APIs to outbreak data. Visualizations implemented upon Web technologies (and that can be as simple as a static image in JPEG, PNG, GIFF, etc.) are by definition first class objects in the Web. What is missing is for the data behind the visualization to also be first class objects on the Web. The way to make that happen is for APIs and Web-native formats. For data on the Web that means JSON and CSV, and to a lesser extent XML (yup, shameless self-plug there).

Epidemic data for the 2014 Ebola Outbreak in West Africa

In February 2014, the first case reports of the West Africa ebola outbreak surfaced in Guinea, although via retroactive tracking it seems the outbreak started in December 2013. For months, outbreak numbers had to be manually screen scraped from Web pages and PDF files. Therefore quality reusable data only became widely available to the public in mid October 2014.

What if someone in Guinea could have started engagingly publishing their own data back in February 2014? What if MSF had been able to cheaply and easily deploy compelling visualizations on their Website? Open data, well defined by a simple Outbreak Time Series Specification, published simply on the Web and visualized by liberally licensed (gratis and libre), pre-built tools would have helped.

It is expected that more and more infectious disease outbreaks will occur in the near future. (via The Guardian)

Map data

As the recent financial support contributed by MSF and the Red Cross to HOT's recent MissingMaps.org project demonstrates, the need for quality maps needs to be solved proactively for future outbreak and other disaster responses. Clearly, HOT is a successful open source community yet in the venture capital fuled software start-up context such open source projects are actually funded by commercial interests. Now with the added fuel, MissingMaps.org's work will proceed much more quickly. But now ask: what tools will global and local parties use that track an outbreak, using the OSM maps?

The vision

Which leads me to the Outbreak Time Series project. Consider the Liberian government's ebola outbreak dashboard ebolainliberia.org

As the site states:

This project commissioned by Liberia's Ministry of Information and Communication seeks to provide a central location for the latest data about the Ebola Epidemic.

This is the sort of thing that should just come out of the box for future epidemics. Note that 15 professionals were involved in producing that site. The goal of this Outbreak Time Series project is for such a team to consist of one amateur "script-kiddie" in the future.

Using Simon Johnson's work as evidence, one D3.js coder can work wonders. And if a data API were available with which to feed data into such visualization, then there is reusable front-end or a global outbreak monitoring network. If such dashboard widgets were made freely available, then in the future a "script kiddie" could put up a dashboard in a few hours. And that would be only the first use of the data standard; next the data would be publishable and aggregate-able which leads to all manner of interesting possibilities...

Outbreak Time Series project blog

http://tigue.com/

Tigue has recorded the ongoing story of the Outbreak Time Series project in his blog, which also is arguably the best place on the Web for pictures of and links to visualizations of the 2014 Ebola 2014 Outbreak in West Africa. Reading through those posts is a great way to see what motivated this project.