Nodalities

From Semantic Web to Web of Data
Nodalities

Subscribe

  • Any Podcatcher
  • Any Feed Reader

Categories

Archives

License

Creative Commons License

Linked Data BOF, WWW2007

I’m now in the Linking Open Data discussion session at WWW2007, which is a great precursor to our Open Data panel tomorrow, and the Linked Data session ‘proper’ on Friday morning…

Given the informal and non-programmed nature of this gathering, there’s a great turnout.

Tim Berners-Lee kicks off, providing some background on Linked Data to contextualise the discussion.

“Unlike Semantic Web data which you keep on your own machine, Linked Data is exciting, it’s viral…”

Chris Bizer, Linking Open Data Project. His presentation is online.

“Don’t talk about the Semantic Web, just do it”.

“Get as much Linked Data as possible online, NOW.”

“Get as many people engaged as possible.”

“Solve the easy problems first, do the hard ones later.”

The project is working to expose some of the existing ‘visible’ data sources such as DBpedia and Musicbrainz, express them in RDF, and look at the practicality of linking them together. They have 600,000,000 RDF triples, comprising over 100,000 connections between various data sets.

One example is DBpedia - extracting structured information from Wikipedia and publishing it openly on the Web. There are examples of the way this works included in Chris’ presentation.

Kingsley Idehen is now at the podium, showing the OpenLink RDF Browser to ‘browse the Data Web’.

Steve Coast (gratuitous plug - on ‘my’ panel tomorrow) from Open StreetMap is asking about the value of converting his large body of existing open data from his own XML format to RDF in order to participate, and kicks off an interesting discussion about value and community efforts. Reaction around the room would suggest that Steve will have plenty of enthusiastic help, should he decide to go down this route.

Giovanni Tummarello from DERI Galway shows Sindice.

Dave Beckett from Yahoo! shows triplr.

Tim shows tabulator, reaching out to previously unknown (to Tim) RDF descriptions such as Bio2RDF, and showing how they can be parsed and interpreted in real time.

All of these tools are clearly aimed at a small set of developers, and have a long way to go before they could be used more broadly. Indeed, most of them would do more harm than good if shown too widely right now…

So clearly a lot of interest and potential around Linked Data… but no clear next step from this session specifically. Unless you’ve got a large amount of data that you want to post on the semantic web for like-minded and skilled developers to play with…

Technorati Tags: , , , , , ,

Leave a Reply