Tuesday, December 20, 2005
Lattice of ontologies
Function/structure descriptions
Side note made to the ONTAC forum
Thus, we claim, the sets of concepts should be defined using controlled vocabularies, without having a certain school of computer science impose theorem proving mechanisms - or allowing that school to make negative evaluations about the controlled vocabularies as being "not expressive". I remind everyone that the theorem proving school has huge power in the peer review processes in the United States, and that literally nothing will be published or funded if they are displeased.
Sir Roger Penrose suggests that they are Emperors without the substance to their arguments that they demand.
But they should be respected IF they develop concept ontologies about theorem proving. But like a child sometimes is in a play pen, they (almost always) want to impose on everyone who would do ontology construction the requirement that the theorem proving ontology be tied into the core of any thing called "an ontology". They have placed themselves into the position of being Emperor.
If we use a better "language" then terms like "semantics", "entailment", "expressive" become more real, in a normal everyday sense. If a description of important foundational work cannot be understood because of special meanings to the terminology used in papers like
http://www.cs.man.ac.uk/~horrocks/Publications/download/2003/HoPa03c.pdf
then the ontology community needs to ask deep questions about what is going on. Is this an esoteric discipline or are we going to provide to the public a new technology so that they can communication normally.