Help us make open aid data accessible…
This week the International Aid Transparency Initiative has received a further boost at the High Level Forum on Aid Effectiveness in Busan, Korea. Announcements from the Canadian government, the Inter-American Development Bank and from Hilary Clinton’s on behalf of the US that they will be joining over 20 other signatories committed to publish detailed and timely aid activity data using a common XML format means that IATI data will soon cover over 75% of Overseas Development Assistance.
With high level commitment to #opendata in the aid sector won, the challenge now continues to build on open data foundations to create prototypes, products and platforms that make a real difference to aid effectiveness. Which is all rather timely for this weekends upcoming ‘open data day‘ and Random Hacks of Kindness events, where data-wranglers and developers will be linking up with designers, statisticians and international development experts to work on digital data projects with impact.
So – what might open data day and Random Hacks of Kindness developers focus on around IATI data? Here are just a few ideas…
IATI Open Data Challenges
Showing IATI on Open Spending
Many IATI data files contain detailed transaction information that could be loaded onto the Open Spending platform developed by the Open Knowledge Foundation, giving access to a range of rich visualisations.
Earlier this week the Guardian covered the use of the Open Spending platform to visualise, for the first time, aid to Uganda, and the national budget, side-by-side. The Uganda visualisation was the result of years of hard work and research, but by creating tools to load IATI open data onto the Open Spending platform, and using Open Spending visualisation tools against IATI data, seeing where aid is being spent, and how it links to country budgets, could be made a lot easier.
Mark Brough from Publish What You Fund has already made a start on scripts to convert IATI Data for use on Open Spending, but there’s lots still to be done. Mark will be at IATI Support site provides a Knowledge Base on working with the data. Mark will be at the London Open Data Day, meeting at the Barbican Centre, to provide pointers to any teams wanting to join in on this challenge.
Aid Data Wiki
Donors use IATI to publish the data they have on aid activities, but often there is other knowledge about an activity out there – some of it as structured data, and other content such as videos, photos, web links and more.
An Aid Data Wiki prototype would explore how IATI aid activity data can be used to bring together additional intelligence and insight around projects – letting users share what they know about aid projects, and providing that information back to to users of IATI data.
I’m hoping we might work on this a bit at Random Hacks of Kindness in Oxford – but I would love to hear from other people interested in exploring this too.
Document Search
The IATI Standard allows donors to list the documents associated with a particular aid activity – but right now there is no easy way to search through all those documents and find the projects they connect to.
We’ve probably outgrown the brief prototype we tried using Google Custom Search, so there’s a challenge to meet in coming up with a good way to search and explore all the documents linked to IATI.
Rich Preview
IATI data is published in individual XML files, usually on a country-by-country basis. Right now, there are only basic preview tools available to look at what a file contains – but there is demand for an easier way to dig into and preview the contents of a file.
Could you create a visually engaging way to explore IATI data files?
Linked IATI Data
We’ve carried out some very brief experiments turning IATI into linked data, and loading it onto the Kasabi platform – but there’s a lot more to be done to make the most of IATI as linked data. There’s great potential to be had from turning IATI into linked data, but some practical challenges still to meet, like:
- Turning the IATI Standard into a draft linked data ontology.
- Creating a sustainable way to convert IATI XML into linked data, and to keep a triple store up to date
- Turning IATI code lists into linked data
- Making links between IATI and other datasets and demonstrating the potential of a linked data approach.
Getting started with IATI data
IATI data is published using a standard XML format which is documented on the IATI Standard website where you can also access the codelists used in IATI data files to browse online or download as data.
Linked to all the available data are kept in a CKAN directory at IATIRegistry.org. The Registry has an API for finding out about datasets.
A third-party platform is provided by Development Gateway at IATIExplorer.org which regularly aggregates together data from all the different publishers in the IATIRegistry and makes it possible to fetch back sub-sets of this data. The IATIExplorer platform runs on top of an XML document store (eXist) which has a RESTful interface for querying the full dataset with xpath, and it can directly run XSLT transformations against the data for you. See the documentation here.
Lots of shared source code for working with the data is available on GitHub and the IATI Support site provides a Knowledge Base on working with the data.
On Open Data Day a number of members of the AidInfo team will be on hand on the #iati IRC channel on Freenode.
