Showing papers on "Knowledge representation and reasoning published in 1978"
TL;DR: A program which understands simple stories about painting and makes heavy use of real world knowledge is described, and there is an extensive discussion of various issues in knowledge representation and how they affect frame representations: modularity, the need for problem solving, worldly vs control knowledge, and cleanliness.
83 citations
01 Jan 1978
TL;DR: The general class of pattern-directed inference systems (PDISs) is defined and its general properties are investigated in this paper, where principles for knowledge representation and system architecture are developed that suggest guidelines for controlling the combinatorial explosion that will accompany efforts to implement intelligent systems with extensive amounts of knowledge.
Abstract: The general class of pattern-directed inference systems (PDISs) is defined and its general properties are investigated. Within a taxonomy of PDISs, the special properties of rule-based systems and two subclasses, production systems and transformation systems, are considered in detail. A comparison between these PDIS properties and contemporary information processing conceptions of human cognition suggests that PDISs provide an excellent basis for cognitive modeling. Principles for knowledge representation and system architecture are developed that suggest guidelines for controlling the combinatorial explosion that will accompany efforts to implement intelligent systems with extensive amounts of knowledge.
68 citations
01 Jan 1978
TL;DR: An extended first order predicate calculus language ‘KS’ is defined as the internal representation language for the deductive question-answering system PLIDIS, and an evaluation graph for the normalised query is constructed.
Abstract: An extended first order predicate calculus language ‘KS’ is defined as the internal representation language for the deductive question-answering system PLIDIS. It serves the functions of a semantic representation language for German, of a knowledge representation language, and of a data base query language. KS incorporates the following extensions: equality, recursively constructed argument terms, ‘list terms’ for representing sets of individuals, and many-sorted domains of individuals. The PLIDIS data base contains ground atomic formulas and axioms. The evaluation of KS-queries proceeds in two steps: first, queries are normalized, i.e., made quantifier-free by means of term-embedding; next, an evaluation graph for the normalised query is constructed. The evaluation of the graph is described.
13 citations
01 Jun 1978
TL;DR: The goal structure of a tutor is outlined, types of conceptual bugs that students have in their understanding of physical processes are described and some of the representational viewpoints necessary to diagnose and correct these bugs are discussed.
Abstract: : Tutorial dialogues can be analyzed as an interaction in which a tutor 'debugs' a student's knowledge representation by diagnosing and correcting conceptual misunderstandings. In this paper some tentative steps toward a theory which describes tutorial interactions are outlined. The paper outlines the goal structure of a tutor, describes types of conceptual bugs that students have in their understanding of physical processes and discusses some of the representational viewpoints necessary to diagnose and correct these bugs.
11 citations
01 Jan 1978
TL;DR: Schank and Schank as mentioned in this paper have made considerable theoretical and experimental advances with respect to the representation of knowledge in memory, and these advances have yielded a much clearer picture of the knowledge structure that is to be established in the instructed subject.
Abstract: The past decade has seen considerable theoretical advances with respect to the representation of knowledge in memory. These advances have yielded a much clearer picture of the knowledge structure that is to be established in the instructed subject (Rumelhart, Lindsay, & Norman, 1972; Schank, 1975a). There have also been advances in understanding the importance of inference procedures, as well as the influence of previous knowledge upon the processing of new information (Abelson, 1975; Minsky, 1968, 1975; Schank, 1975b). However, these insights and advances are fragmentary in several important respects, particularly with regard to advancing instructional theory. Although a reasonably clear understanding of knowledge representation has emerged, there have been few theoretical and experimental advances in the understanding of knowledge acquisition and utilization. There is little that offers promise towards a theory of cognitive learning.
9 citations
01 Jan 1978
TL;DR: The motivation for this research has been to discover techniques for organizing information that will optimize learning.
Abstract: The acquisition of knowledge through reading text is a common source of learning in an instructional setting. In a typical learning environment, a student must study texts containing information on related topics and situations, and then integrate that information into a coherent knowledge representation. As any educator knows, the knowledge actually acquired by a student in this situation is only a small subset of all information to be learned. This paper investigates the influence of the structure of presented textual information on the learning of related information. In essence, the following question is addressed: When sets of topically related material containing shared knowledge are to be learned, how are they best presented to the learner? The motivation for this research has been to discover techniques for organizing information that will optimize learning.
7 citations
04 Dec 1978
TL;DR: This work investigates how the user himself can introduce his own expert knowledge into his data base system through rapid language extension so that it may then respond intelligently to his curt queries and commands.
Abstract: A major thrust of artificial intelligence research is how to build knowledge of the application domain into computer systems. We investigate how the user himself can introduce his own expert knowledge into his data base system through rapid language extension so that it may then respond intelligently to his curt queries and commands.Illustrations of rapid language extension using the REL System are presented and discussed.
6 citations
05 Nov 1978
TL;DR: A system of computer programs, MEDIKAS, that interacts with an expert in order to acquire information on an area of medical knowledge and manifests a level of intelligence in its discourse with the expert by utilizing inference on information already acquired.
Abstract: Renewed interest in the use of computer-based system for medical problem solving has directed attention to the problems involved in representing medical knowledge in a form that will enable such knowledge to be 1) acquired efficiently, with guidance from the automated system, 2) maintained in such a form that modifications to it can be made easily, and 3) utilized by one or more systems of machine-based logic or inference. Our work in studying knowledge representation and devising structures for the interactive acquisition and efficient storage of information in knowledge bases has been directed toward these goals. A system of computer programs, MEDIKAS has been created. The system interacts with an expert in order to acquire information on an area of medical knowledge. In addition to operating interactively, the system manifests a level of intelligence in its discourse with the expert: by utilizing inference on information already acquired, it deduces relationships when possible, and thus promotes both the efficiency of the expert and the consistency of the content.
5 citations
01 Sep 1978
TL;DR: BAOBAB is concerned with the problem of parsing - recognizing natural language sentences and encoding them into MYClN''s internal representation, and uses a semantic grammar in which the non-terminal symbols denote semantic categories.
Abstract: Until a recent knowledge-based system is able to learn by itself, it must acquire new knowledge and new heuristics from human experts. This is traditionally done with the aid of a computer programmer acting as intermediary. The direct transfer of knowledge from an expert to the system requires a natural-language processor capable of handling a substantial subset of English. The development of such a natural-language processor is a long-term goal of automating knowledge acquisition; facilitating the interface between the expert and the system is a first step toward this goal. This paper descrtbes BAOBAB, a program designed and implemented for MYCIN (Shortliffe 1974), a medical consultation system for infectious disease diagnosis and therapy selection. BAOBAB is concerned with the problem of parsing - recognizing natural language sentences and encoding them into MYClN''s internal representation. For this purpose, it uses a semantic grammar in which the non-terminal symbols denote semantic categories (e.g., infections and symptoms), or conceptual categorles whlch are common tools of knowledge representation in artificial intelligence (e.g., attributes, objects, values and predicate functions). This differs from a syntactic grammar in which non-terminal symbols are syntactic elements such as nouns or verbs.
5 citations
01 Jan 1978
TL;DR: The paper describes the design of a knowledge manipulation system, currently under construction, programmed in a LISP-like language which incorporates parallelism, and allows programs to be run in remote environments.
Abstract: Knowledge manipulation tasks are computationally intensive : one way to provide sufficient computing power is to use a loosely coupled network of microprocessors. For this approach to be efficient it is necessary that the overall task be divisible into parts which can be run in parallel and which are, to a degree, self-contained. It is argued that contemporary approaches to knowledge manipulation systems fulfil these criteria. The paper describes the design of such a system, currently under construction. It is programmed in a LISP-like language which incorporates parallelism, and allows programs to be run in remote environments. Aspects of this language, its implementation and the widerlying hardware are described.
01 Dec 1978
TL;DR: This Report is a collection of papers presented at the "Workshop on Natural Language for Interaction with Data Bases" held at IIASA in Laxenburg, Austria from January 10 to 14, 1977.
Abstract: This Report is a collection of papers presented at the "Workshop on Natural Language for Interaction with Data Bases" held at IIASA in Laxenburg, Austria from January 10 to 14, 1977.
The papers describe the research and results in attempting to produce a viable, useful, and flexible interface to various systems in Europe (acronymically AQL, PLIDIS, USL, DONAU, DILOS, KAIFAS, etc.) and in North America (OWL, INGRES, and LIFER). Most of these interfaces present to the user the feeling of working in an environment of relatively free and forgiving syntax. This is in marked contrast to the rather rigid syntax required by most commercial data base systems in their natural (mostly English) query languages. In addition there are discussions of the categorization of the semantic relationships within some natural languages as an aid to both understanding and knowledge representation.
01 Sep 1978
TL;DR: Investigation into some of the newer technologies to determine if they have useful applications to the problem of automating the fusion of multisource data are initiated, including natural language processing, production rules, and pattern recognition.
Abstract: : This report initiates investigations into some of the newer technologies to determine if they have useful applications to the problem of automating the fusion of multisource data. The main technology areas discussed are natural language processing (NLP), production rules, and pattern recognition. The theory of possibilities is also considered. An integrated data fusion system is postulated which would employ a number of different, interacting techniques. Early stages of system processing would require NLP techniques to restructure narrative data into a form useful for automated fusion with other kinds of data. The use of 'frames' as knowledge representation structures is considered for this part of the processing. Techniques involved in later processing could include production rules and pattern recognition. Possible applications of these two techniques to data fusion problems are discussed. (Author)
13 Apr 1978
TL;DR: A system, called SARA, is described that implements some of the recent techniques used to organize the computers view of the world and an ability to understand events is demonstrated by generating a paraphrase of a short input story.
Abstract: Recent development in natural language processing have shown that the inclusion of world knowledge in a processing system will inhance the understanding capabilities of the computer. This paper describes a system, called SARA, that implements some of the recent techniques used to organize the computers view of the world. An ability to understand events is demonstrated by generating a paraphrase of a short input story. The systems limitations are discussed as a function of the unsolved problems in designing a universal scheme for an adaquate knowledge representation.
18 Jul 1978
TL;DR: In this article, the design of a knowledge manipulation system based on a loosely coupled network of microprocessors is described. But this approach is not suitable for large-scale knowledge manipulation tasks, as the overall task is divided into parts which can be run in parallel and which are, to a degree, self-contained.
Abstract: Knowledge manipulation tasks are computationally intensive : one way to provide sufficient computing power is to use a loosely coupled network of microprocessors. For this approach to be efficient it is necessary that the overall task be divisible into parts which can be run in parallel and which are, to a degree, self-contained. It is argued that contemporary approaches to knowledge manipulation systems fulfil these criteria. The paper describes the design of such a system, currently under construction. It is programmed in a LISP-like language which incorporates parallelism, and allows programs to be run in remote environments. Aspects of this language, its implementation and the widerlying hardware are described.