Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!tut.cis.ohio-state.edu!medulla.cis.ohio-state.edu!jj From: jj@medulla.cis.ohio-state.edu (John Josephson) Newsgroups: comp.ai Subject: abduction workshop Message-ID: Date: 23 Feb 91 18:28:13 GMT Sender: news@tut.cis.ohio-state.edu Organization: Ohio State Computer Science Lines: 81 * ! * * ! * * ! * * ! * * ! * * ! * * ! * * ! * * ! * REMINDER * ! * * ! * * ! * * ! * * ! * * ! * * ! * * ! * * ! * AAAI 1991 Workshop Towards Domain-Independent Strategies for Abduction to be held at the 9th National Conference on AI (one day during July 14-19, 1991) Upcoming deadline: Submissions should be sent so as to arrive by March 1, 1991. If you are anticipating any problem with this deadline, please get in touch with Venu Dasigi by e-mail or phone. All potential participants should submit a short (less than 1-page) summary of their work and a list of relevant publications. (Easy! You can send by e-mail.) Those wishing to discuss their work at the workshop should submit extended abstracts of papers (up to approximately 4 single-spaced pages); abstracts of work in progress are encouraged. Send 5 copies to: Venu Dasigi, Department of Computer Science, Wright State University Research Center, 3171 Research Boulevard, Dayton, OH 45420. E-mail: vdasigi@cs.wright.edu Phone: 513-259-1395 or 513-873-3201. Acceptance notices will be sent April 15, 1991. Here again is the Call for Participation. Call for Participation The workshop will focus on applications of abductive inference with the objective of discerning which computa- tional strategies have what degree of domain dependence. Two types of work will be of interest in the workshop: that which examines abductive processing in more than one domain, and that which describes a particular application of abduction from an engineering perspective, with an eye to what can be generalized and applied to other domains. Treatments of other aspects of abduction, such as its characterizations, models, causality, explanatory nature, etc. are encouraged insofar as they address the main focus. Of interest are issues of identifying, character- izing or clarifying procedural (information processing strategy) aspects of abduction. That is, there appears to be consensus that medical diagnosis or determining struc- tures of chemical compounds are (logically) abductive problems, but it is not clear whether MYCIN or DENDRAL should considered to be (procedurally) abductive systems. It is also hoped that this workshop might serve to relate existing theory to applications and implementations. Most of the applications of abduction that have appeared in the literature so far are in diagnosis, plan recognition and natural language processing. It appears that other applications, notably in speech processing, vision (e.g., object recognition) and even plan genera- tion, etc. are possible and are being currently investi- gated. One of the aims of this workshop is to encourage participants to share their preliminary investigations in such areas, and discuss what interesting issues underlie them, and whether most such issues are common to different applications. The schedule also includes a keynote talk by Harry Pople. Organizing Committee: Michael Coombs (mcoombs@nmsu.edu) Venu Dasigi (vdasigi@cs.wright.edu) - Correspondent Jerry Hobbs (hobbs@ai.sri.com) John Josephson (jj@cis.ohio-state.edu) - Chair Yun Peng (ypeng@algol.cs.umbc.edu) Brought to you by Super Global Mega Corp .com