Why linked data is cool

I recently spent 30 minutes with one of my colleagues, Michael Panzer, learning about the joys of linked data and why it is so cool. He has created a linked data proto-type for Dewey (http://dewey.info ) that is now publicly available for re-use. Michael told me the three rules for linked data which I’ll roughly paraphrase as:

1. Use URIs for everything
2. Every URI gives a useful description of the object
3. In the description provide more links to other useful stuff

I must say I got pretty excited from my layman’s low level technical view of the mashability of this type of data service. If like me you are new to linked data I would explain it simply as:

· You have a data element that is unambiguous, such as Dewey number 641 and you know this number is indeed a Dewey classification number.
· You know the site where someone has created a linked data service that will give you additional information about that number
· You can now programmatically create the URL from your piece of data that will retrieve additional information about the data object which you can re-use in your own service.

Here is one example of a URL for Dewey 641: http://dewey.info/class/641/about. The 641 was data in your system and the rest of the URL was created by your program.

Note that this actually returns you what both 640 and 641 are classifications for: Home & family management and Food & drink respectively. But what is even nicer is it returns that information in nine languages appropriately tagged so you can re-use it in any of those languages. Another nice feature is while the machine sees the raw data a human can see HTML.

Pretty nifty service for easily beefing up your own data with further information. Michael has also blogged about this new service at: http://tiny.cc/qu4UV