This is the first blog post by Sarah, who has has been working with us on some copywriting, marketing and sales activities. Some of Sarah’s work has recently featured on our new website.

Linked Data rocks. It’s the ideal technology for representing data on the web. However that message is spreading relatively slowly and it’s not being used as much as it could be. As Linked Data fans, what can we do about it?

It all starts and ends with the user of the data. If we know what data they want, and what they want to do with it, then we can make sure we provide data access in a way that suits them.

But who are they? Current Linked Data users split roughly into four types:

  1. the information seeker: someone who just wants to know about a specific topic and who like visualisations and tables;
  2. the researcher: an analyst who wants access to the raw data but prefers to use Excel and other mainstream tools;
  3. the web developer: a programmer who wants to create their own visualisations or apps and generally likes JSON APIs;
  4. the Linked Data expert: who wants a SPARQL endpoint so they can really get to the bones of the data.

The last group is small on numbers when compared to the others but they are the group that a lot of current Linked Data is aimed at. Many Linked Data publishing initiatives have a few words of introduction and a text box for a SPARQL query, which is grand for the experts but not so great for everyone else. And even these RDF loving experts still need something more from data providers so they can get up to speed with the data quickly. Often it’s this group, and the web developer group, who use the raw data to create something understandable and accessible for the other groups.

Developers deserve a good user experience too. Yes, they have programming skills, they may be well used to working with data and some of them may even be SPARQL experts (though in our experience most aren’t). But they need to spend time understanding the structure of data, how it is interconnected, the data modeling patterns that have been employed and so on. Examples that they can copy, paste and adapt are a big help in getting started.

Anything the data publisher can do to make that process quicker and easier will pay itself back many times in the lifetime of the data. It is a good investment for those who understand the data in detail to provide as much useful information as possible, rather than requiring each individual user to do that work themselves.

So what’s the way forward?

  • Good documentation - both about the data and about ways of accessing it.
  • Follow best practices and standard patterns - because it makes things more predictable for data users.
  • Recognise that there are different types of audience for the data and try to meet the needs of all of them.
  • Connect different views of the data so people can follow links from one view to another.

The strength of Linked Data is that it is easy to generate slices of data (possibly combining multiple sources) in a variety of formats. We should exploit that strength to offer the data to different audiences in the formats they prefer. That’s our philosophy and we’re currently working on enhancing the data presentation features in PublishMyData to meet that goal.

Keep up to date with our news by signing up to our newsletter.
Thanks for reading all the way to the end!
We'd love it if you shared this article.