Since the subject has been mentioned, I'd like to point to our investigation of the semantics of neural networks using category theory. It started much more recently than Andree's, of which I became aware only a year or two ago. We differ somewhat in approach, although we've been using some of Andree's ideas. My partner Tom Caudell and I have been doing research in neural networks for many years, our approach being a combination of mathematics and cognitive neuroscience. We started applying category theory in our semantic modeling a few years ago and are now developing architectures from it. Some references are easily accessible on my web page at http://cialab.ee.washington.edu . I'm under Affiliate Faculty, and there's a link to publications near the bottom of my home page. The categorical ones are conference papers, toward the bottom of the publications page (though if you're that interested, please let me send you a corrected version of the 2000 Como paper). We welcome feedback as well as questions. Mike =========================================================================== e Michael J. Healy A FA ----------> GA mjhealy@u.washington.edu | | | | 13544 23rd Place NE Ff | | Gf Seattle, WA 98125 | | USA \|/ \|/ ' ' FB ----------> GB e "I'm a natural man." B ============================================================================