Friends of Semantic Compositions

January 2009

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31

Site Statistics

Blog powered by Typepad

« Mystery solved | Main | This is for you, LOTR fans »

February 06, 2004



"That's why, as nice as the semantic web idea sounds, SC doesn't think it will work until somebody says "enough!", and takes control of what documents mean -- an event nobody is particularly hoping will happen."

A lurking word, "Microsoft", just shuffled its feet in the dark alleys of my mind. I shuddered.

Semantic Compositions

This isn't going to do much for your confidence, Virge:

Microsoft MindNet


We'll have an automatically updated repository of "world knowledge" force-fitted to Microsoft's ontology and frequently distorted by googlebombing practices.
I suppose this shouldn't be too disheartening. Right now we have an information network our grandparents would have thought of as utopian -- and the biggest single use for it is porn.

"I've seen the future, baby: it is murder" - Leonard Cohen


I'd suggest that the SemWeb approach implicitly acknowledges that a single, global "world knowledge" is not likely to happen (and probably isn't desirable!). The RDF and OWL languages do make it relatively straightforward to make 'local' ontologies and mappings between them (they are defined globally, but their application may be much narrower). There are certainly big issues on the global scale - trust etc. But most of the time in practice it isn't necessary to deal with every ontology under the sun, just the ones relevant to the immediate task.

I think it goes along the lines of the Alan Kay quote : "Simple things should be simple; complex things should be possible". The current web gives us lots of 'simple' (-ish) things, Semantic Web technologies may not offer all the answers just yet, but at least they offer a chance of complex things being possible.

The comments to this entry are closed.