Results


Report 83-20.pdf

Classics (Collection 2)

Jeffrey S. Rosenschein and Vineet Singh Heuristic Programming Project Computer Science Department Stanford University Stanford, CA 94305 Abstract Meta-level control, in an Artificial Intelligence system, can provide increased capabilities This improvement, however, is achieved at the cost of the meta-level effort itself. This paper outlines a formalization of the costs involved in choosing between independent problem-solving methods: the cost of meta-level control is explicitly included. It is often desirable for Artificial Intelligence systems to make use of explicit knowledge about what they know; this tneta-level knowledge allows a program to direct its own activities in an informed and efficient manner [I] [21. The use of meta-level knowledge by a system to control its own actions is called'new-level confrol. If we are to gain efficiene; thi-migh the use of meta-level effort, \'.e must be sure that %'.hat is aved at the base level is not canceled by what is expended at the rnota-level.


Report 77-33.pdf

Classics (Collection 2)

Reprint of a paper appearing in: Proceedings AWL Conference, SIGART/SIC;PLAN Combined issue, August, 1977. We suggest that the concept of a strategy can profitably oe viewLd as know,,tie about /ow to st,'t ct porn aII1011g a set of plausibly useful knowledge sowces, and explore the framewoi k foi knowledge organization which this implies. Meta rules are also considered in the broader context of a tool for programming. We show that they can be conciciered a medium for expressing the criteria for.etri7val The utility of this as a prugrairiming mechanism is considered.


S Report 77 16 Stanford KSL

Classics (Collection 2)

Meta-Level Knowledge: Overview and Applications Randall Davis and Bruce C. Buchanan Computer Science Department Stanford University Stanford, California 94305 Abstract The representation and use of knowledge has been a central problem in Al research. A range of different encoding techniques have been developed, along with a number of approaches to applying knowledge. Most of the effort to date however, has concentrated on representing and manipulating knowledge about a specific domain of application, like game-playing, natural language understanding, speech understanding, etc. It begins by defining the term, then explores a few of its varieties and considers the range of capabilities it makes possible. Four specific examples of meta-level knowledge are described - as they have been implemented in a program called TEIRESIAS.