Nodalities

From Semantic Web to Web of Data
Nodalities

Subscribe

  • Any Podcatcher
  • Any Feed Reader

Updates

Follow us on:

Categories

Archives

License

Creative Commons License

Garlik releases open source triple store, 4Store

4storeGarlik CEO Tom Ilube is increasingly coming to represent a voice of reason in the UK’s ongoing angst about Identity, with many a hysterically gibbering Home Office official put in their place by Tom’s more reasoned words in debates on the Today programme and across the UK’s mainstream media.

As the company’s press materials note,

“Garlik, the online identity expert, was founded by Mike Harris, founding CEO of Egg plc, former Egg CIO Tom Ilube and former British Computer Society president Professor Nigel Shadbolt. As the first company to develop a web-scale commercial application of semantic technology, Garlik enables consumers to protect themselves against identity theft and financial fraud.”

According to Wikipedia, ‘Egg… is now the world’s largest internet bank,’ so effective management of identity information is clearly nothing new to Ilube and his team.

Founded in 2005, Garlik has secured some £4.5million from 3i, Doughty Hanson and Noble Venture Finance to offer products such as their DataPatrol solution for tracking sensitive personal information online, and the less ’serious’ measure of online status, QDOS.

Behind the scenes, data is aggregated from across the open Web and various proprietary databases, and stored in Garlik’s own RDF triple store.

Now the company is releasing their triple store — 4store — under a GNU GPL license and making it available for download. Capable of scaling to handle as many as 60billion triples (perhaps at least three times more than their closest competitors), 4store has the potential to address many concerns about the scalability of triple store technology.

I took the opportunity to talk with Garlik’s Tom Ilube and 4store’s designer, Steve Harris, before the launch and the result has just been released as a podcast.

 
 Standard Podcast [36:09m]: Play Now | Play in Popup | Download (413)

This conversation was recorded on Tuesday 14 July, 2009.

One Response

  1. Atanas Kiryakov Says:

    This is a nice development and it is good for the market RDF databases!

    It would have been even more interesting if there was any public justification about the scalability and performance claims made in this post and elsewhere. For instance, I personally believe that our store can lodd 100B triples at 400 KSt/sec given the hardware used by some of the competitors. But our practice is not to throw figures in the public space before we can justify them and provide sufficient evidence and technical details.