hunter-larry@YALE.ARPA (04/08/88)
From: Larry Hunter <hunter-larry@YALE.ARPA> Hello - I'm new to this list, and I'd like to introduce myself. I'm in the process of finishing my Ph.D. in artificial intelligence at Yale, under the direction of Roger Schank. My thesis work is a computerized information retrieval and consultation system in lung tumor pathology called IVY. The system gradually gains expertise by discovering "paradigm cases" in its experiences. By paradigm case, I mean not necessarily a prototypical instance, but a case that embodies some particular prescription for accomplishing a useful task (e.g. it is helpful in distinguishing between two commonly confused phenomena). The program finds paradigm cases by a process of knowledge acquisition planning: It develops knowledge goals (things it would like to know about) based on problems it had previously, and then it can figure out what aspects of its knowledge or of future experiences might be useful in finding out that information. An experience that can be used to satisfy a knowledge goal is saved as a paradigm case. The program is quite efficient in that it can find many useful cases without an exhaustive search. I believe that I can apply this technology to the domain of genetics and molecular biology, generating programs that can plan for and achieve particular knowledge acquisition goals by making inferences and looking up information from multiple databases. The knowledge acquisition planning process is quite flexible, so other compuational methods for making inferences from biological databases can easily be integrated into it. Brief discussions I have had with Christian Burks at LANL, Clay Saunders at the Human Gene Mapping Library and Larry Kingsland at the National Library of Medicine have all been very encouraging. I'd now like to ask for some advice. I'm thinking about where I should go after I get my degree. Where are the significant projects in computational biology? What are their research interests and inclinations? Who is interested in hiring AI people with some biological background? I've been talking to Stanford and the National Library of Medicine, but there have to be other projects out there. I'm particularly interested in assistant professor positions, but I'm open to other possibilities. Anybody have any suggestions? Thanks! Larry Hunter HUNTER@YALE.EDU -------