06 April 2008

Semantic Web, RDF and URI in the real world

My interest in Semantic Web has slowly diminished as so few applications and data are available. But I keep an eye on it and from time to time, I read an article.
This time : Cool URIs for the Semantic Web a note from the W3C. It is about the way you should serve content and uses URI for RDF, HTML and "Real Things". Conclusion is : there are no good way to do it, they present 2 very different solutions among others. So since both solutions are far from perfect, every one is making his semantic web.

One strange point : "a naming scheme should not confuse things and the documents representing them.". I have a mixed feeling about this. It states that a "Thing" should have an URI, but that URI should be distinct from the URI of the document describing the "Thing". So what is the point of having an URI if it just redirect to a real document. Just for reference ? Then why should it return something at all ? It is more practical. Sure. Then it looks like a mess resulting from theoretical and practical views.

Among the articles I read and didn't blog about :
- Some editors/publishers may see a value in adding metadata to their content (articles, reviews ...). This would provide more content to be retrieved with semantic Web applications.
- Computer will be able to add more significant metadata automatically. Normal users would just have to validate.
- Data-portability, will raise the need to detach the data and replace the format (HTML/CSS) with metadata. So one will be able to move his data from one site to another, each one providing its own format and applications.

No doubt that we are going to benefit from the semantic web on a large scale one day, but not today !
Post a Comment

October summary - 2017-10

10 October Science Looks like the p-value has been under fire for years. https://www.nature.com/news/one-size-fits-all-threshold-for-p-value...