Daniel’s Blog

Serialising Semantics

December 10th, 2007 by daniel

OWL (Web Ontology Language) is a Semantic Web vocabulary. With OWL we can build up terms and relationships of terms to give explicit meaning to those terms, this obviously has a benefit when we come to actually giving instances to these terms. (just so we are on the same wavelength: Watch is a term, whereas myWatch is an instance)

There are other ways to build meaning and ontologies which are in the Semantic Web, such as: OIL, DAML+OIL, RDFS, SKOS. The all allow you to establish varying amounts of depth of semantics, but probably not as much (or easily) as in OWL.

Ontology Languages in the Semantic Web are based on a formalism called Descriptions Logic, which is quite an interesting area but I won’t get into this very complex field here (it involves some mathematical symbols and an understanding of first order logic).

My question I pose to the reader (mainly if you have any AI, Semantic Web or Descriptions Logic Knowledge):

Are there any ways which you can think of to “serialise” semantics, and doing so in a way that isn’t already a Semantic Web formalism?

By serialise (serialize in American), I actually mean actually writing this in a format which can be manipulated/processed on a computer/machine.

Now this is probably going to be a symbolic system (i.e. based on logic symbols), but not necessarily. I say not necessarily because the answer may well be a more connectionist one (by which I mean based on technologies such as Artificial Neural Networks, Genetic Algorithms/Programming), but I doubt it, as symbolic systems tend to be better at analysing language.

So I would like to turn this to you because I would like your thoughts, and maybe we can discuss.

Technorati Tags: semanticweb, ontologies, semantics, meaning

No Comments

<!--<div id="semanticbrowsing" style="text-align:center;"> OpenLink RDF Browser 1 OpenLink RDF Browser 2 <area shape="rect" coords="242,5,346,102"