Ebola Outbreak Data on the Web Overview - JohnTigue/idots GitHub Wiki

This page is a curated collection of ebola outbreak data on the Web. This collection provides an overview of the state of the situation via notable examples. It is not meant to be an exhaustive catalog.

Recommendations for Data to Use in Web Apps is the place to go if looking for data to code against.

The examples are broken down into three types.

1. Data catalogs: human readable documents and Web sites which enumerate collections of known URLs at which ebola outbreak numbers are available.
2. Data repositories: Web sites at which multiple ebola datasets can be found. These are Web apps where data resides, as opposed to items in #1 above which are simply contain pointers to where data can be found.
3. Data via APIs: ebola outbreak data available via private APIs.

The goal of this EbolaMapper project is to have open source code (Apache 2.0 licensed) which can interactively visualize ebola outbreak data in Web browsers (and do the same during future epidemic outbreaks, ebola or otherwise). By definition, folks are free to do whatever they like with Apache 2.0 licensed code. For example, say, a national health ministry wishes to visualize their private info: they can deploy EbolaMapper privately and knock themselves out, it's no one else's business.

Ebola outbreak data is available at this project via the Outbreak Time Series Specification, licensed as open data under PDDL. Real success will be when other sites also provide data through the same APIs.

The following information was collected early in development while searching for dataset that could be format via the Outbreak Time Series Specification such that EbolaMapper could have some data to visualize.

1. Gapminder's list
   List of datasources on the Web (a Google Doc)
   They also maintain a calendar of report release dates

2. Ebola Data Jam
   http://eboladata.org/dataset
   Analysis can be found on this wiki's page, Datasets Listed on eboladata.org

3. data.gov
   https://catalog.data.gov/dataset?q=ebola
   No analysis of this search result yet but a quick glance leads one to think there is significant overlap with what the eboladata.org effort has curated.

4. Another GoogleDoc spreadsheet
   A spreadsheet of ebola outbreak data
   This one was found via http://www.reddit.com/r/ebola

These repositories currently have ebola outbreak data but not real data rich usable APIs, yet, although that is changing. These two are the most important, as I believe they will be the two first place to deploy the Outbreak Time Series Specification due to their current activity in the ebola response.

1. The Humanitarian Data Exchange (HDX)
   See HDX repository

2. Open Humanitarian Data Repository (OHDR)
   See OHDR repository

There are a very few, about four, instances of Ebola2014 data available in any form that could be called an API. CSV is being excluded from this definition, even though its inclusion could be argued, especially in light of standards like #HXL. No use of #HXL for Ebola2014 data is known of.

Since before this project there was no API for outbreak data, the instances of ebola-outbreak-data-in-JSON that have been found have simply been JSON quickly banged out for the purposes of getting the data from one server to one client. Let us call this a "private API." These bits of JSON served their one-time purposes.

A goal of this project is that ebola outbreak information be available via a standard APIs. This is what is being called Outbreak Time Series Specification. For future outbreaks, ebola or otherwise, having a simple yet well defined API will enable that speedy proliferation of outbreak data across the Web and into Web site applications and mobile device apps (read: iPhone and Android apps).

This project is the first and, as of yet, only source of HTTP Resources which contains representations of outbreak data made available via the Outbreak Time Series Specification.

See JSON Ebola2014 Data Found on the Web for detailed analysis of the few private APIs found on the Web. They were used during analysis to come up with a global harmonized non-ebola specific set of Outbreak Time Series Specification, which has turned out to be one of the two major deliverables of this project.