2005-01-18

How many "person" concepts in the Semantic Web?

Need to define "Person"in your local ontology? Unless you want to add yet another identifier for this concept, your contribution to the Semantic Mess, go ask Swoogle Ontology Dictionary and pick up your choice. At first ranks come with no surprise classes from FOAF http://xmlns.com/foaf/0.1/Person and WordNet http://xmlns.com/wordnet/1.6/Person.
But there are about 400 other SW resources, classes and properties using the same name. Food for thought ... Why so many already? What about re-usability? What about aggregation of data and federation of knowledge? (see previous post) How many of those resources declare equivalence with other ones? And how many are actually used?

2 comments:

  1. Person: "...400 ... classes and properties using the same name..." is an excellent observation. It supports my feeling that we need spend far more time on solving the problems of meaning coordination and negotiation, establishment of common knowledge, and the automated evolution and acquistion of language. We have less need for support for First Order Logic, SQL like query facilities, or other features at this time. Those problems have well defined solutions, automating shared interpretations of symbols by machines does not.

    ReplyDelete
  2. I suspect that us software folks are simply terrible at being human in the path of logic and computer systems. "Meaning", ie. semantics, is left as an exercise to Arts Major folks rather than being an integral part of our core.

    More semantics to the IT people! Shove epistemology down their throats before it's too late and we're stuck with *thousands* of class:Person.

    Sigh.

    ReplyDelete

Comments welcome