The Department of Computer Science & Engineering
cse@buffalo

CSE725: Implementing KR&R Systems, Fall, 2000

Prof. Stuart C. Shapiro

Registration No. 384478

Thursdays, 3:00 - 5:00, 224 Bell Hall


This seminar will be a group project. We will review the implementation of SNePS 2.5 and the partial implementation of SNePS 3.0, and complete the implementation of SNePS 3.0.

SNePS is a propositional semantic network knowledge representation and reasoning system. The basic SNePS data structure is a fully-indexed network database. SNePS supports inference by path-following in the network, and also by a general natural-deduction-style combined forward and backward chaining system that uses a parallel, eager-beaver control system with memoization. An assumption-based belief revision system is an integral part of SNePS inference, and an acting component is integrated with the reasoning component. Basic SNePS operations include unification, subsumption, and operations on sets. A graphical user interface is available, as well as a standard text-based interface. SNePS 3.0 uses the ANALOG logic, which uses a kind of generalized restricted quantifier syntax. SNePS 3.0 is partially implemented in Common Lisp and CLOS. We will either continue that implementation, or switch to Java, or do both.

Required background of participants: ability to program in Common Lisp and/or Java; knowledge of Artificial Intelligence at least at the level of CS572, including familiarity with knowledge representation, logic, and reasoning systems. People with stronger backgrounds than the minimal are particularly welcome. People with backgrounds and interest in AI reasoning systems, people with backgrounds and interest in implementing logic programming languages, and people with background and interests in GUIs, especially graph drawing, are welcome.

Participants can expect to significantly improve their knowledge and abilities in all the topics mentioned in the previous two paragraphs.

All participants will be expected to participate actively in seminar meetings, to contribute to the implementation of SNePS 3.0, and to lead discussions on their part of the system.

If you intend to register for this seminar, you must discuss that intention with Prof. Shapiro before the start of the Fall semester, and obtain his permission to register.

It would help if before the Fall semester starts, participants would read:

  1. Sections 1 - 3 of Stuart C. Shapiro and William J. Rapaport. The SNePS family. Computers & Mathematics with Applications, 23 2-5, (January--March 1992), 243-275.

  2. Syed S. Ali and Stuart C. Shapiro. Natural language processing using a propositional semantic network with structured variables. Minds and Machines, 3, 4, (November 1993), 421-451.

  3. Stuart C. Shapiro, SNePS: A Logic for Natural Language Understanding and Commonsense Reasoning. In Lucja Iwanska & Stuart C. Shapiro, Eds., Natural Language Processing and Knowledge Representation: Language for Knowledge and Knowledge for Language, AAAI Press/The MIT Press, Menlo Park, CA, 2000, in press.

  4. Stuart C. Shapiro, An Introduction to SNePS 3. In Bernhard Ganter & Guy W. Mineau, Eds. Conceptual Structures: Logical, Linguistic, and Computational Issues. Lecture Notes in Artificial Intelligence 1867. Springer-Verlag, Berlin, 2000, 510--524.

Stuart C. Shapiro <shapiro@cse.buffalo.edu>