Enter the hashtags — A story of user centred design and the Humanitarian Exchange Language (HXL)
The humanitarian exchange language (HXL) is a data standard for humanitarian responders which fits on a postcard. It helps improve information sharing during a humanitarian crisis
It’s been a year since HXL reached v1.0 as shown above. However, on the journey to this point, HXL went through a couple of iterations each with a very different approach. Quite often I get questions from technically focused people as to why the current form was chosen (as it has a lot of technical limitations), and whether changes can be made to improve it.
So here is the HXL journey to help explain why HXL is what it is.
2011 The First attempt — Technical Perfection
The first attempt at HXL was the ideal technical solution. It was based around the principle of linked data or semantic data. The general principle is that each data category is uniquely defined at an URI and relationships are defined for this data category. HXL used the particular flavour called RDF triples. An example of a RDF triple is:
The components of a triple, such as the statement “The sky has the colour blue”, consist of a subject (“the sky”), a predicate (“has the colour”), and an object (“blue”).
The concepts and relationships were defined for humanitarian context. Below is a subset of some of the data relationships.
Confused?
A lot of people were. Enough in fact that there was very low adoption. Imagine you are working in a disaster context, where the most widely used tool is spreadsheet software and information managers and analysts build data templates on the fly in high pressured environments. They are very unlikely to spend the time to learn how to convert their data into RDF triplets even if the tools exist.
Below is my summary of the situation
Other viewpoints have been suggested for why the first attempt didn’t move forward.
It was never intended for data creators and managers to use and understand RDF, but rather for them to use tools on top of it, including a HXLator which turned spreadsheet data to RDF. RDF may not have been the fundamental problem, but rather talking about RDF before the accompanying tools existed. Although the first HXLator was tested by small groups, it was never really pushed for adoption.
2014 The Second Attempt — Enter the Hashtag
Due to the low adoption rate the approach was revised in 2014. Enter the HXL hashtags. The HXL standard now consists of entering agreed upon hashtags in the row below the headings to add context. Here is the 30 second tutorial:
This approach is cooperative rather than competitive. It doesn’t attempt to fundamentally change the way the data creators and managers work. They still operate in their tool of choice, the standard is in the spreadsheet format, it is lightweight enough to be squeezed on a postcard and the basic concepts communicated in 30 seconds.
And most important of all, people are using it and adding the HXL hashtags to their data.
As someone on the coding side of HXL it certainly means I have to work harder and think more creatively for the solutions, but the big plus is, I am implementing systems with real data, for real purposes rather than admiring a technical solution’s theoretical potential. The solution with the users’ context at heart was the solution that more people adopted.
To find out more about HXL visit the HXL standard webpage.
Browse HXLated data on Humanitarian Data Exchange.
Read about how we are using HXL at the British Red Cross.
