This article presents an experiment in knowledge-intensive programming within a general problemsolv:ng production-system architecture called Soar In Soar, knowledge is encoded within a set of problem spaces.

From attributes 8 3 Implementation 8 3.1 Overview of Meta-Rulegen 3.2 Algorithm 10 3.2.1 Approach from object rule 11 3.2.2 Approach from attributes 14 4 Preliminary Results 15 5 Conclusion 17 ENHANCING PERFORMANCE OF EXPERT SYSTEMS BY AUTOMATED DISCOVERY OF META-RULES Abstract Machine learning can be used to formulate new meta-level knowledge. A small MYCIN-like medical diagnosis system was constructed as a starting point. Two heuristic methods are used in a program called Meta-Rulegen to form meta-rules from the knowledge base in the diagnosis system. In a preliminary study, 63 meta-rules were formed automatically and, by judiciously selecting a set of meta-rules, the efficiency of the diagnosis system can be improved significantly without degrading the quality of advice. This study suggests that meta-rules can be learned automatically to improve the efficiency of rule-based systems.

Corona is a prototype language for describing designs. The goal is to capture all the information created in the process of designing a component. The information about a design includes the specification of its structure and behavior. It is important to organize this knowledge hierarchically, similar to the way a designer does in refining a design incrementally. Also, we would like to capture the process of the design in addition to its specification.

The single unifying assumption in this work is that one or more of the interacting agents will be using artificial intelligence techniques to guide their actions (including, of course, their communication actions). We call this the "intelligentagent paradigm." Within this broad categorization, the many individual efforts to give Al systems the capability to interact with other rational systems are seen as potentially increasing efficiency (by harnessing multiple reasoners to solve problems in parallel) or as necessitated by the distributed nature of the problem (e.g., distributed air traffic control

Although informal models of evidential reasoning have been successfully app'ied in automated reasoning systems, it is generally difficult to define the range of their applicability In addition, they hay., not provided a basis for coherent management of evidence bearing on hypotheses that are related hierarchically. The Dempster-Shafer (D-S) theory of evidence is appealing because it does suggest a coherent approach for dealing with such relationships However, the theory's complexity and potential for computational inefficiency have tended to discourage its use in reasoning systems In this paper we describe the central elements of the D-S theory, basing our exposition on simple examples drawn from the field of medicine. We then demonstrate the relevance of the 0-S theory to a familiar expert system domain, namely the bacterial organism identification problem that lies at the heart of the MYCIN system. Finally, we present a new adaptation of the D-S approach that achieves computational efficiency while permitting the management of evidential reasoning.within

I. Introduction [he HELIOS system is a tool for conducting research into the design.

We have developed an automated system for the design of laboratory experiments in molecular biology. The system uses a planning method known as skeletal plan refinement that attempts to emulate the human cognitive task of experiment design. This paper describes the theory, history, and implementation of the design system and illustrates its function in the domain of DNA cloning experiments.

Stanford -- KSL that Think. Consideration of the of the new 16-bit integrated circuits that phenomenal progress of the past 30 years leaves one with a feeling of have allowed computers oi small size and considerable power to be developed. The only certainty in sight is that scientists. BRUCE G. BUCHANAN is Professor of In addition to game playing early Al work focused on techniques for solving Computer Science Research at Stanford small symbolic reasoning problems. Researchers continue to ponder these problems (Overleat) Illustration by f red Nelson as well.

The ideas themselves have been implemented on Stanford's SAIL system.

This means that we by watching the expert diagnose a patient. Our approach relies heavily on a close correspondence are trying to create a framework whereby an between the system and a human expert problem solver's knowledge organization with respect to knowledge organization, inference and knowledge acquisition methods are modeled methods and discourse language. The described system is a major component of a learning as similarly as possible to human problem by watching system being created to facilitate solvers.