This is a primary area of research to my organization as
syntax searches across multiple knowledge bases result in too many hits, just
like Google. We are using OWL-DL to build the semantic search and tying in term
lists for the object relationship to truly get a much more finite set of
results. This of course means having to build specific ontologies in OWL and
building term lists. And for each knowledge base, with its different format
tagging, may require additions to the literature specification.
However, I see large grass roots effort to publish these RDF
and OWL ontologies to make it easier for people to search. There may already be
RDF and OWL literature specifications for PDFS, Word, Word Perfect, HTML, XML,
etc that will allow organizations to effectively text mine the corpus of
knowledge. I can also picture Google hiring SMEs and ontologies to create these
general ontologies so that they can more intelligently index the web.
Shoebottom, BA, MA2
Tel: (506) 674-5439
Cell: (506) 476-4023