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

« SC caves in | Main | Maybe determiners aren't so predictable »

February 05, 2004

Comments

Danny

"The basic idea of the semantic web is that making inferences about the meaning of free text is hard."

No it isn't, nor is it about juggling on a unicycle being hard!

The basic idea is that adding explicit resource descriptions to the web is pretty easy and potentially very useful, and drawing inferences from these isn't too difficult.

Even if reading natural language *was* easy for machines, some form of processing would be needed to make use of it, something perhaps like formal-logic based approach behind Semantic Web technologies.

It's also worth noting that the kind of descriptions expressed in RDF/OWL needn't have any connection to natural language at all - e.g. a digital camera can timestamp photos, this can be encoded in RDF/XML and reasoned about/queried in Semantic Web systems.

The comments to this entry are closed.