The use of explicit goals for knowledge to guide inference and learning

Abstract

Combinatorial explosion of inferences has always been a central problem in artificial intelligence. Although the inferences that can be drawn from a reasoner's knowledge and from available inputs is very large (potentially infinite), the inferential resources available to any reasoning system are limited. With limited inferential capacity and very many potential inferences, reasoners must somehow control the process of inference.Not all inferences are equally useful to a given reasoning system. Any reasoning system that has goals (or any form of a utility function) and acts based on its beliefs indirectly assigns utility to its beliefs. Given limits on the process of inference, and variation in the utility of inferences, it is clear that a reasoner ought to draw the inferences that will be most valuable to it. This paper presents an approach to this problem that makes the utility of a (potential) belief an explicit part of the inference process. The method is to generate explicit desires for knowledge. The question of focus of attention is thereby transformed into two related problems: How can explicit desires for knowledge be used to control inference and facilitate resource-constrained goal pursuit in general? and, Where do these desires for knowledge come from? We present a theory of knowledge goals, or desires for knowledge, and their use in the processes of understanding and learning. The theory is illustrated using two case studies, a natural language understanding program that learns by reading novel or unusual newspaper stories, and a differential diagnosis program that improves its accuracy with experience.

Untitled

“…Any such rule must make a statement about the goals of the program, not just about the content of the domain. A similar argument can be made for the use of knowledge goals to focus the inferencing process for understanding, explanation, or diagnosis (Ram, 1990d;Ram & Leake, 1991;Ram & Hunter, 1992).…”

“…Other questions arise from other kinds of gaps in the system's knowledge or other kinds of difficulties during processing. Our theory of questions and knowledge goals is discussed in Ram (1989;1991) and Ram and Hunter (1992). For the purposes of this article, questions may be thought of as identified gaps in the system's memory or knowledge base, representing what the system needs to know for the purposes of the reasoning tasks that it is carrying out.…”

Goal-Directed Metacontrol for Integrated Procedure Learning

“…While most of these systems focus on a uniform learning method, our work supports a wider range of learning methods with different strengths. Ram and Hunter (1992) introduce the notion of explicit knowledge goals to capture gaps in the system's knowledge, defining knowledge goals as comprising both the specification of the missing knowledge and the task enabled by it. In addition, they propose augmenting knowledge goals with a utility measure to help drive the inference process.…”

A Functional Theory of Creative Reading

“…No reader, for example, would have the time, memory, and other resources to read every single word of a story (even a short story) and consider the possible extrapolations and extensions resulting from each. The reader's current environment, knowledge, goals and tasks (Ram & Hunter, 1992), and cognitive resources (e.g., Just & Carpenter, 1992) interact to control the level of understanding which is attempted. Therefore, a comprehensive theory of the reading process must also explain when and why in-depth reading occurs, as well as when and why shallow reading, such as skimming or skipping text, occurs.…”