APIcon UK: Why Developers Should Take Linked Data Seriously

If you aren't familiar with linked data or the Semantic Web, it's probably a good time to get better acquainted with the concepts, as it seems to be a popular theme that is rippling through the Web development world, with promises of exciting possibilities. At this year's APIcon in London, Markus Lanthaler, inventor of Hydra and one of the core designers of JSON-LD, talked about the importance of optimizing data architecture for an integrated future, focusing on linked data as a promising solution.


Markus Lanthaler speaking at APIcon UK

So what is the Semantic Web? David Berlind, editor-in-chief of ProgrammableWeb, gives a detailed explanation in a recent post, but to put it in a nutshell, as it is stated on Wikipedia,

According to the W3C, "The Semantic Web provides a common Framework that allows data to be shared and reused across application, enterprise, and community boundaries." The term was coined by Tim Berners-Lee for a web of data that can be processed by machines.

Why is it relevant? Because it essentially means that users will be able to find, share and combine information more easily.

In his address, Lanthaler illustrated how the use of multiple APIs could be simplified by introducing one central standard to work by, as opposed to using a whole bunch of different languages. It's about one standard and one SDK. Linked data is built on the same principle that the Web is built on, and instead of using natural language, it uses definitions — subject, predicate and object — all identified with URLs (which can be looked up if not recognized). So, it's about making it possible for a machine to make the connections between things like a human would, ultimately providing the end user with a much better experience that requires far less time and effort.

Lanthaler described linked data as a tool that enables you to create compelling websites without having to create all of that data yourself. By giving a bit more context and understanding to a machine, you can extend what is possible. He went on to talk about ways that developers could implement these principles effectively, specifically in the case of Web APIs. First, he touched on JSON-LD, a user-friendly solution to the restrictions JSON presents in handling linked data. JSON-LD makes it possible to serialize this data in a JSON-based format.

He then went on to talk about Hydra, which is a core vocabulary that allows developers to create completely generic API consoles, SDKs and so on. He highlighted BBC Things as an example of how this is currently used. BBC Things uses Semantic Web technologies that allow full access to all BBC data and is built on its linked data Platform. This makes it possible for the system to make intuitive connections across products and content that would not have been possible to achieve without a whole lot of manual effort.

One of the slides shown in the session featured a quote from Evan Sandhaus of The New York Times that sums it all up quite nicely:

What [linked data] will let you do on the back end is pretty revolutionary. It lets you answer questions, not that you couldn't answer before, but [for which] it would have been way too hard to collect, sanitize and curate the data.

For more information on Hydra and JSON-LD, you can visit the Hydra website and even join the Hydra W3C Community Group to get more involved.

Be sure to read the next APIcon article: APIcon UK: Why You Should Consider Your API a Product