Web content standard plays antics with semantics

The semantic development of internet content has dominated the World Wide Web Consortium (W3C), with European researchers putting forward a standardisation proposal that hopes to supersede XML.

The proposal consists of solutions for the description of linguistic content which can be processed in such a way that a machine can understand it.

Professor Dieter Fensel from Freie Universität Amsterdam said: "Our ontologies are even more abstract [than XML]. They represent a unified level on which different XML descriptions can be projected."

Fensel, together with Ian Harrocks from the University of Manchester, is one of the authors of the standardisation proposal DAML + OIL, or darpa agent mark-up language and ontology instance language.

XML describes in general ways how, for example, a book ordering page looks structurally - author, price, publisher and so forth - while ontology represents formal descriptions of the meaning system of our language.

So, for example, linguists or computer scientists break the words 'warm' or 'southern France' down into individual atoms of meaning - 'what the word implies to the recipient'.

More objective elements, such as details of the temperature scale, are then added to these more subjective factors. In this way the computer does not process character strings alone, but rather interprets character strings.

Companies such as Semantic Edge or Empolis, both based in Germany, are already implementing such computer linguistic research into products, particularly search engine applications.

By Jürgen Höfling, reporter, http://www.silicon.de