I edited this post, but it seems I forgot to save what I edited. It is not 1.5 but 1.7 millions (check the link) and according to Mr. Finin:

“To answer these questions, we first have to decide what we are counting. While Swoogle has found over 1.7M Semantic Web documents and more than 70K of these define some terms, we consider fewer than 1% to be ontologies. Moreover, we can observe considerable variation in the way ontologies are defined and evolve. Many ontologies are constructed and maintained collaboratively, like a Wiki. Design tradeoffs also vary, with some large comprehensive ontologies like CYC and other small and modest vocabularies like FOAF.”

It seems that only 1% (70k) of the documents are ontologies. Okay, I said “only” but 70k ontologies is far too much in my humble opition.

However, the SW will certainly not start with 70k different ontologies, but probably with few, specific and well defined ontologies, like DC, FOAF, SIOC, DOAP, etc. (there I am talking about the adoption of the SW for general, day-to-day applications, and not for specific fields like medicine, genomic, etc.

But personally, I think we reached a first step with RSS content. In 6 years or so, it started from nothing and now hundred of RSS readers have been developed, most of the big web sites export their data using it. So I think that it could be done: people will eventually understand why they should care about exporting their data using ontologies. Six years ago, people asked the same question: why should I export my data in RSS? Now they know: they want to crawled by Technorati, they want to build a readership, etc.

What is good with RSS (and other syndication format) is that it teached to people why they should care about freely exporting their content in a format easily processable by computers. Now, we have to extent this to other ontologies.

Thanks for commenting this blog post!

Take care,

Salutations,

Fred