Monday, March 2, 2009

Semantic Web and Linked Data: what does this all mean

This is a topic that a lot of people are discussing:

http://www.semanticsincorporated.com/2009/02/the-siloed-view-of-the-semantic-web-as-linked-data.html

http://www.web3beat.com/2009/02/the-definitional-challenges-of.html

http://tomheath.com/blog/2009/03/linked-data-web-of-data-semantic-web-wtf

First things first. We are all up to putting names on things and making things even more confusing: Semantic Web, Web of Data, Web 3.0, Giant Global Graph. We all have the same objective: make the web's data more valuable, interoperable and accessible.

We can make smart applications once we have raw data that can be understood and accessed easily by machines.

The W3C has worked on a series of recommendations and technologies: RDF, OWL, SPARQL, RDFa, GRDDL, etc. All in order to fulfill TBL's original Semantic Web vision. In my opinion, a Semantic Web application is one that uses these technologies that have been standardized.

However, this is not the only way! I agree with you. Jinni.com, Pandora.com etc do awesome "smart" stuff with movie and music data respectively. These are definetely semantic applications. However they do not use the specific W3C technologies. So following my definition, they are not Semantic Web applications, only semantic applications.

Linked Data is NOT the Semantic Web. I agree. It is an important outcome of the Semantic Web effort to put the Semantic Web technologies (RDF, OWL, etc) in practice! You can't do "smart stuff" with only linked data. You need to apply more AI techniques. However, when the data is linked, you will be able to provide the Serendipity and Discovery that the Web 3.0 promises (ouch.. not sure if I should say that, however here is my definition of the Web 3.0 http://joshdilworth.com/post/57811660/web-3-0-and-knowability-diagram)

No comments: