In January this year DFID began to publish details of every aid project in the International Aid Transparency Initiative (IATI) Standard XML schema. There are now a set of XML downloads referenced from the IATI Registry covering countries, regions and multilateral aid activities, all delivered through a RESTful API, and refreshed every month with up-to-date data. DFID also publish a more human-readable version of this information on the Projects Database on the DFID website.

The Hewlett Foundation are the second organisation to publish in IATI format, and others are following on close behind.

Eating my own dogfood

Owen Barder blogged recently that organisations publishing data should be forced to use that data on their own websites – so that “eating their own dogfood” would drive up quality. As I don’t get the chance to do this at work, I’ve been having a play with the IATI data in my own time. Just for fun, you know.

Here are a few initial prototypes. I will blog implementation details in the next few posts, and include code.

Charting using amCharts

I first saw the amCharts charting tool in ForeignAssistance.gov, the US aid portal. The tool takes XML or CSV data and generate Flash or Javascript charts – pie, column/bar, line, bubble and so on. The tool is very straightforward to set up, and data can be set up using XSL transformation of IATI XML data.

These sort of charts would be useful to:

• get an overview of what a donor is doing in a country
• use as a widget on a site dedicated to aid in that country
• (if aggregated across donors in a country), provide an overview of what activities are considered important
• show past and future aid flows for a country

There are still some challenges – for example aggregating the sectors (health, infrastructure, water) rather than the very granular sectors listed.

See with the charts in action. Use 2-letter ISO codes to look up a country you’re interested in.

Yahoo! Pipes

Yahoo! Pipes is a tool that enables manipulation of internet data feeds using a set of pre-determined tools. This pipe retrieves the country file for, in this case, Uganda (UG), filters for currently active projects, and produces a list of aid activities matching those criteria.

Note: the IATI schema needs to have the XML directive added at the top to enable it to be read by Yahoo! Pipes.

These tools would be useful to:

• aggregate information for a country from different sources (for example through the IATI Registry)
• mashup using connections between aid projects and other indicators
• transform or filter the data without using XSL

Go on, have a play.

WordPress Country View

In the WordHack at last years WordCamp UK, the WordHack team developed a prototype to import aid data into a WordPress custom posts structure. As the WordHack was based on non-IATI XML, I’ve adapted this for IATI data.

In short, openaid.org.uk allows an import of IATI XML data for a particular country into a WordPress instance, creating a new Custom Post from each IATI activity (aid project), and dealing with DFID’s hierarchical project/component structure. The code regularly checks the data source for updates and creates new versions accordingly.

This could be used by a Civil Society Organisation (CSO) in a developing country to:

• aggregate all donor activities for their country,
• allow citizens to add comments on each donor activity using the WordPress commenting capability, and
• keep track of the changing nature of the projects through regular updates.

Here’s the example site with a very few test projects.

Next steps

These are just a few initial ideas and prototypes. Please add your ideas below, or, if you’re really keen, visit the new IATI Support forum and send in your ideas.

Disclosure: Although I work for DFID, these prototypes were developed in my own time and using my own equipment. Yes, I’m a geek.

The WordHack at WordCamp UK in Manchester this year was really exciting. By 4.30 pm on day 2 the team had created a stunning WordPress site that:

• Reads and transforms XML data on DFID aid projects into a format ready for import into WordPress
• Imports the XML into a WordPress Custom Post Type for each DFID aid project, creating new post versions if the source data has changed
• Links comments to the post version, so that someone can view comments in the context of the data at the time
• Displays the data in a clean user interface, focussing on the conversation that people can have around aid projects

User Story
The concept (User Story) of the site is that people (either UK citizens or citizens of developing countries) can view summary details of DFID aid projects, and can start conversations around individual projects. Ultimately, this may be best used for sub-sets of the full 3000-odd projects based around communities of interest, for example by a community within a developing country.

Why WordPress?
The new Custom Post Types in WordPress provide the framework needed to handle more structured data around aid project information. Other core features are the commenting capability, use of categories and tags for classification, and the built-in post versioning system. These would require a significant effort to write from scratch.

Credit to the Team
I was really pleased with what the team achieved, and it builds on work done at the recent Aid Information Challenge. It clearly shows what can be done with government information when the data is provided in a reusable format.

The keys to the success were:

• Shaun Hare‘s organisation was critical. He set up the WordHack wiki page, canvassed for ideas early, and allowed for discussion within the group.
• Getting ideas early really helped, as people could develop the ideas over a couple of weeks, and come to WordHack with a better-formed solution in mind. It also helped to break up the problem into a number of logical components that enabled everyone to do something useful.
• The skill and dedication of the team was phenomenal. Each person chose a different component and delivered something useful, and Chris and Shaun tied it all together on Sunday afternoon.

More details
The technical details of the site are available on the OpenAid project page on the WordCampUK Wiki, and we hope to have a Google Code project set up soon.

Next steps
The next steps are likely to be:

• Add Ajaxy goodness to the front page to allow the most recent conversations to magically appear, Twitter style.
• Make the import plugin generic to allow any XML data to be transformed into the Custom Post Type – possibly by providing a URL for the XML and an XSL transform.
• Create a map based on country data to show where the conversation is happening.
• and many more…

If you are interested in continuing involvement then please follow @openaiduk on Twitter.