Making Links at the BBC

Ian and I spent last Friday at BBC Television Centre in London. For anyone of my generation who grew up in the UK this place probably has an almost mythical status, as the place to send your competition entries or milk bottle tops for the latest Blue Peter appeal. We were there for a workshop on the theme of the Semantic Web, organised by Nicholas Humfrey and Patrick Sinclair from BBC Audio and Music Interactive.

Not only was it a privilege to get a look inside this great institution, it was great to see so many BBC people turn up to hear about the Semantic Web. Nick and Patrick had put together a very nicely structured programme, introducing people to the Semantic Web from the conceptual level of Linked Data (that was my bit), through a talk on DBpedia by Georgi Kobilarov, to the highs and lows of enterprise scale RDF storage as revealed by Steve Harris of Garlik, and finally to interfaces for structured data as presented by Daniel Smith from the University of Southampton. Hope all the slides will be linked to from the BBC Radio Labs blog in due course. In the meantime you can find mine here.

Aside from the inherent pleasure associated with talking to people about the Semantic Web, the highlight of the day for me was getting a sneak preview of the Linked Data work that’s going on within the BBC, and will hopefully soon see the light of day on the public Web. The /programmes area of the BBC site will be home to large amounts of RDF data about programmes going out across all channels, and each will be identified by a dereferenceable URI.

This is a huge deal, and testament to the hard work put in by people like Nick, Patrick and Michael Smethhurst from the BBC, with input from people like Yves. There is already a public commitment to linked data principles at http://www.bbc.co.uk/programmes/developers , but what impressed me most was the extent to which linking to external data sets seems to be baked into the thinking from day one. Expect to see strong links to Musicbrainz in the first instance, and no doubt to many more data sets over time.

The BBC are well ahead of the game here. They don’t have an angry mob of license-fee payers at the gates demanding access to BBC data in RDF, with chants of “give us our data, we’ve paid for it already” (or hopefully something more poetic). This mob will never materialise. They’ve seen the willingness of the BBC in this area with previous initiatives such as the Catalogue, and are down the pub dreaming up ways to use this data. With the advent of the current work on Linked Data and /programmes the non-mob have even more to dream about.

Perhaps as a publicly-funded organisation the BBC is obliged (morally or otherwise) to be a good citizen of the Web of Data. However, I don’t get the impression that that’s what this is about, in the first instance at least. I’m left with the feeling that this is a result of a bunch of guys really getting the Web of Data, and seeing the value that links can bring to their organisation.

Comments are currently closed.