Can You Explain Linked Data?
The other day I was asked, or perhaps challenged, by the ODI's Amanda Smith to come up with a one sentence description of Linked Data that didn't use any acronyms or technical terms. I came up with this:
Linked Data uses the concepts and technologies of the Web to describe the world
which seemed to do the trick (original tweet).
It saddens me that such a powerful and, at its heart, simple technology attracts so much negativity. Actually it doesn't just sadden me, it irritates me, to put it mildly. I put it less mildly in my SemTechBiz Keynote in August. TimBL's original definition is:
- Use URIs as names for things
- Use HTTP URIs so that people can look up those names.
- When someone looks up a URI, provide useful information, using the standards (RDF*, SPARQL)
- Include links to other URIs so that they can discover more things.
Antibodies are raised in some people to every line of this. URIs are too long; you can't do real time look ups in an app, RDF is way too complicated and who understands that SPAQRL shit and follow your nose crap anyway? Just use JSON, man
I exaggerate — but only a little.
SPARQL isn't hard. Thinking in graphs is hard — seeing and keeping a mental picture of all the links and nodes is hard, especially when your data describes something that isn't trivially simple. So, yes, it takes effort and it takes an understanding of what it is you're modelling. But if you need to to model something as complex as the real world, it's a very powerful tool.
The challenge is putting that power in the hands of people who have antibodies racing round their bloodstream. JSON-LD is the big lever we have here but there's more to do, more bridges to build. And that's what my focus will be on in 2015. Who knows, we might even see an end to scepticism on this scale. Well, I'm allowed a Christmas wish aren't I?