In this paper we will be concerned with such reasoning in its most general form, that is, in inferences that are defeasible: given more information, we may retract them. The purpose of this paper is to introduce a form of non-monotonic inference based on the notion of a partial model of the world. We take partial models to reflect our partial knowledge of the true state of affairs. We then define non-monotonic inference as the process of filling in unknown parts of the model with conjectures: statements that could turn out to be false, given more complete knowledge. To take a standard example from default reasoning: since most birds can fly, if Tweety is a bird it is reasonable to assume that she can fly, at least in the absence of any information to the contrary. We thus have some justification for filling in our partial picture of the world with this conjecture. If our knowledge includes the fact that Tweety is an ostrich, then no such justification exists, and the conjecture must be retracted.

The particular task in question is the interpretation of the electrical signals generated by the heart muscle, known as the electrocardiographic (Eco) interpretation.

The traditional approach involving protracted interaction between a knowledge engineer and a domain expert is viable only to the extent that both these resources are available; this approach will not meet the apparently exponential growth in demand for expert systems. A solution to this dilemma requires rethinking the way knowledge-based products are built. An example of this reappraisal of methodology appears in Michie (1983), and is based on the principle of formalizing and refining the knowledge implicit in collections of examples or data bases. Dietterich and Michalski (1983) give an overview of methods for learning from examples. There are many such, all based on the idea of inductive generalization. One of the simplest of these methods dates back to work by Hunt in the late fifties (Hunt et al., 1966). Each given example, described by measuring certain fixed properties, belongs to a known class and the'learning' takes the form of developing a classification rule that can then be applied to new objects. Simple though it may be, derivatives of this method have achieved useful results; Kononenko et al. (1984), for example, have managed to generate five medical diagnosis systems with minimal reference to diagnosticians.

An increasing number of Artificial Intelligence (Al) programs are implemented on high-performance workstations with a bitmap display, a mouse input device, and a keyboard. The programming environment (usually a dialect of LISP) generally provides support for multiple, overlapping windows, and various kinds of menus including pop up menus. The user can move, reshape, close, and scroll the windows. Additionally, a programmer can designate arbitrary regions of a window to be selectable with the mouse. This means that a user can invoke an action by pressing and releasing a mouse button while the mouse cursor is in the designated region.

Current subprojects include applications of the program Gerring.

These paradigms include 1) clinical algorithms, 2) statistical analysis of collections of patient data, 3) mathematical models of physical processes, 4) decision analysis.

This article reviews the strengths and limitations of five major paradigms of medical computer-assisted decision making (CAOM): 1) clinical algorithms.

I cis is the final report for Contract N-00014-79-C-03C2, covering the period of 15 March 1979 through 14 March 1982. The goal of the project was to develop methods for representing teaching and problem-solving knowledge in computer-based tutorial systems. One focus of the work was formulation of principles for managing a case method tutorial dialogue; the other major focus was investigation of the use of a production rule representation for the subject material of tutorial program. The main theme pursued by this research is that representing teaching and problemsolving knowledge separately and explicitly enhances the ability to build, modify and test complex tutorial programs. Two major corr Jter programs were constructed.

This papei describes a device-independent diagnostic program called DART. DART differs from previous approaches to diagnosis taken in the Artificial Intelligence community in that it works directly from design descriptions rather than MYCIN-like symptom-fault rules. DART differs from previous approaches to diagnosis taken in the design automation community in that is more general and in many cases more efficient. DART uses a device-independent language for describing devices and a device-independent inference procedure for diagnosis. The resulting generality allows it to be applied to a wide class of devices ranging from digital logic to nuclear reactors. Although this generality engenders some computational overhead on small problems, it facilitates the use of multiple design descriptions and thereby makes possible combinatoric savings that more than offsets this overhead on problems of realistic size.

Reprinted from IJCAI, August 24-28, 1981, Vancouver, British Columbia. Used by permission of the International Joint Conference on Artificial Intelligence, Inc.; copies of the Proceedings are available from Morgan Kaufmann Publishers, Inc., 95 First Street, Los Altos, CA 94022, USA. DART: An Expert System for Computer Fault Diagnosis James S. Bennett Heuristic Programming Project, Computer Science Deparunent Stanford University, Stanford, CA 94305 Clifford R. Hollander IBM Scientific Center, 1530 Page Mill Road Palo Alto, CA 94304 To appear in International Joint Conference on Artificial Intelligence, August 24-28, 1981, Vancouver, British Columbia. A. Intro Juction We describe an application of artificial intelligence techniques to computer system fault diagnosis, in particular, we have implemented an automated consultant that advises IBM field service personnel on the diagnosis of faults occurring in computer installations. The consultant identifies specific system components (both hardware and software) likely to be responsible for an observed fault and offers a brief explanation of the major factors and evidence supporting these indictments.