[topicmapmail] Re: [topicmaps-comment] Genetic PSIs [Re: Topi c Map domain, paradigmatic PSIs ...]

[Martin Bryan]

Jim

> I take John Sowa's advice to heart. He likes (paraphrasing his words)
broad,
> shallow, very generic ontologies, as opposed to the SUMO and Cyc people
who
> are trying to build comprehensive, deep, detailed ontologies.

I like the sentiment, but its wrong in one sense. Ontologies should never be
"very generic": they should be highly targeted. They need to be shallow but
because the relationships between items change for different scenarios an
ontology should not try to cover the whole universe of discourse but only a
clearly defined part of the universe. For example, how useful would an
ontology that broadly covered the subject of food be for the identification
of data resources at Oak Ridge?

Martin