scispace - formally typeset
Search or ask a question

Showing papers on "Domain knowledge published in 1983"


Journal ArticleDOI
TL;DR: In this article, the authors propose a framework for expert system design, in particular about how the domain knowledge may be decomposed into substructures, and corresponding to each is a separate substructure specializing in that type of problem-solving.
Abstract: Our group's work in medical decision making has led us to formulate a framework for expert system design, in particular about how the domain knowledge may be decomposed into substructures. We propose that there exist different problem-solving types, i.e., uses of knowledge, and corresponding to each is a separate substructure specializing in that type of problem-solving. Each substructure is in turn further decomposed into a hierarchy of specialist which differ from each other not in the type of problem-solving, but in the conceptual content of their knowledge; e.g.; one of them may specialize in "heart disease," while another may do so in "liver," though both of them are doing the same type of problem solving. Thus ultimately all the knowledge in the system is distributed among problem-solvers which know how to use that knowledge. This is in contrast to the currently dominant expert system paradigm which proposes a common knowledge base accessed by knowledge-free problem-solvers of various kinds. In our framework there is no distinction between knowledge bases and problem-solvers: each knowledge source is a problem-solver. We have so far had occasion to deal with three generic problem-solving types in expert clinical reasoning: diagnosis (classification), data retrieval and organization, and reasoning about consequences of actions. In novice, these expert structures are often incomplete, and other knowledge structures and learning processes are needed to construct and complete them.

271 citations


Journal ArticleDOI
TL;DR: This paper will present the basis for this view of expertise, the reasoning model it implies, and a computer program which begins to implement the theory, called SHRINK, which models psychiatrie diagnosis and treatment.
Abstract: Two major factors seem to distinguish novices from experts. First, experts generally know more about their domain. Second, experts are better than novices at applying and using that knowledge effectively. Within AI, the traditional approach to expertise has concentrated on the first difference. Thus, “expert systems” research has revolved around extracting the rules experts use and developing problem solving methodologies for dealing with those rules. Unlike these systems, human experts are able to introspect about their knowledge and learn from past experience. It is this view of expertise, based on the second distinguishing feature above, that we are exploring. Such a view requires a reasoning model based on organization of experience in a long-term memory, and incremental learning and refinement of both reasoning processes and domain knowledge. This paper will present the basis for this view, the reasoning model it implies, and a computer program which begins to implement the theory. The program, called SHRINK, models psychiatrie diagnosis and treatment.

225 citations


Journal ArticleDOI
Janice S. Aikins1
TL;DR: A system that uses a representation of prototypical knowledge to guide computer consultations, and to focus the application of production rules used to represent inferential knowledge in the domain is presented.

191 citations


Journal ArticleDOI
TL;DR: SeeK as discussed by the authors is a system that provides a unified design framework for building and empirically verifying an expert system knowledge base using case experience, in the form of stored cases with known conclusions, to interactively guide the expert in refining the rules of a model.

144 citations


Proceedings Article
22 Aug 1983
TL;DR: A poorly designed knowledge base can be as cryptic as an arbitrary program and just as difficult to maintain Representing control knowledge abstractly, separately from domain facts and relations, makes the design more transparent and explainable.
Abstract: A poorly designed knowledge base can be as cryptic as an arbitrary program and just as difficult to maintain Representing control knowledge abstractly, separately from domain facts and relations, makes the design more transparent and explainable A body of abstract control knowledge provides a generic framework for constructing knowledge bases for related problems in other domains and also provides a useful starting point for studying the nature of strategies

113 citations


Book ChapterDOI
01 Jan 1983
TL;DR: This chapter is a case study in the mechanical mapping of domain-specific problems onto general methods, using as a detailed example the derivation of a heuristic search procedure for the advice “avoid taking points.”
Abstract: A key problem in learning by being told is operationalization: the development of procedures to implement advice that is not directly executable by the learner, such as the advice “avoid taking points” in the card game hearts. One way to operationalize such advice is to reformulate it in terms of a general “weak method”, such as heuristic search. This chapter is a case study in the mechanical mapping of domain-specific problems onto general methods, using as a detailed example the derivation of a heuristic search procedure for the advice “avoid taking points.” The derivation consists of a series of problem transformations leading from the advice statement to an executable procedure. The operators used to perform these transformations are implemented in a program called FOO as domain-independent transformation rules that access a knowledge base of task domain concepts. Some of the rules construct a crude generate-andtest procedure; others improve it by deriving new heuristics based on domain knowledge and problem analysis. To test its generality, FOO was also used to operationalize a music composition task; many of the same rules proved applicable.

111 citations


Proceedings Article
08 Aug 1983
TL;DR: A formal, but pragmatic, method of recording and organizing human expertise into a knowledge-based system is presented and experience gained from testing and from expert feedback is described.
Abstract: A formal, but pragmatic, method of recording and organizing human expertise into a knowledge-based system is presented. Practical considerations and methods which increase system validity while minimizing demands on human domain specialists are explored. The methodology concentrates on domain definition (background knowledge, references, situations, and procedures), on fundamental knowledge formulation (elementary rules, beliefs, and expectations), and on basal knowledge consolidation (review and correction cycles). Experience gained from testing and from expert feedback is described.

91 citations


Proceedings Article
08 Aug 1983
TL;DR: This paper demonstrates a methodology for collecting and analysing observations of experts at work, in order to find the conceptual framework used for the particular domain, and develops a representation for qualitative knowledge of the structure and behavior of a mechanism.
Abstract: The ability to identify and represent the knowledge that a human expert has about a particular domain is a key method in the creation of expert computer system. The first part of this paper demonstrates a methodology for collecting and analysing observations of experts at work, in order to find the conceptual framework used for the particular domain. The second part develops a representation for qualitative knowledge of the structure and behavior of a mechanism. The qualitative simulation, or envisionment, process is given a qualitative structural description of a mechanism and some initialization information, and produces a detailed description of the mechanism's behavior. This "vertical" slice of the construction of a cognitive model demonstrator, an effective knowledge acquisition method for the purpose of determining the structure of the representation itself, not simply the content of the knowledge to be encoded in that representation. Most importantly, it demonstrates the interaction among constraints derived from the textbook knowledge of the domain, from observations of the human expert, and from the computational requirements of successful performance.

67 citations


Journal ArticleDOI
TL;DR: There is a section on instructional strategies that an instructional designer or teacher can use to help optimize the learner's use of the seven kinds of prior knowledge for acquiring, organizing, and retrieving new knowledge.
Abstract: Any comprehensive theory of instruction must include ways to optimize the acquisition, organization, and retrieval of new knowledge. An important concern in this regard is making new knowledge meaningful by relating it to prior knowledge. Although meaningfulness is usually thought of in terms of relating new knowledge to prior superordinate knowledge (as with the advance organizer), there are at least six other kinds of prior knowledge that can facilitate the acquisition, organization, and retrieval of new knowledge. Seven kinds of prior knowledge are described below, followed by a section on instructional strategies that an instructional designer or teacher can use to help optimize the learner's use of the seven kinds of prior knowledge for acquiring, organizing, and retrieving new knowledge.

62 citations



Journal ArticleDOI
TL;DR: Concepts of “knowledge engineering” that are most relevant in designing and building knowledge-based decision support systems are discussed.

Journal ArticleDOI
TL;DR: This special issue introduces the importance, diversity, and vigor of knowledge representation as a research activity to a wider audience by mapping out the basic approaches to knowledge representation that have developed over the years.
Abstract: In contrast to conventional database systems, AI systems require a knowledge base with diverse kinds of knowledge. These include, but are not limited to, knowledge about objects, knowledge about processes, and hard-to-represent commonsense knowledge about goals, motivation, causality, time, actions, etc. Attempts to represent this breadth of knowledge raise many questions: (1) How do we structure the explicit knowledge in a knowledge base? (2) How do we encode rules for manipulating a knowledge base's explicit knowledge to infer knowledge contained implicitly within the knowledge base? (3) When do we undertake and how do we control such inferences? (4) How do we formally specify the semantics of a knowledge base? (5) How do we deal with incomplete knowledge? (6) How do we extract the knowledge of an expert to initially "stock" the knowledge base? (7) How do we automatically acquire new knowledge as time goes on so that the knowledge base can be kept current? This special issue introduces this important area of artificial intelligence to a wider audience. The core of the 15 articles, contributed by a broad spectrum of researchers on various aspects of knowledge representation, show the importance, diversity, and vigor of knowledge representation as a research activity. This introduction provides some background and context to these articles by mapping out the basic approaches to knowledge representation that have developed over the years.

Journal ArticleDOI
TL;DR: It was found that the usual adult superiority in speed of processing could be markedly reduced if children possessed equivalent amounts of domain knowledge and this effect was domain specific.

Proceedings Article
08 Aug 1983
TL;DR: An extension of FRL (Frame Representation Language) which supports the encoding of reasoning knowledge within a frame-based formalism is described, called HPRL (Heuristic Programming and Representation language).
Abstract: This paper describes an extension of FRL (Frame Representation Language) which supports the encoding of reasoning knowledge within a frame-based formalism. The extension is called HPRL (Heuristic Programming and Representation Language). The declarative representation of reasoning knowledge in the same formalism that is used to represent domain knowledge results in a powerful tool for the construction of expert systems. Reasoning knowledge is easy to describe, examine and modify. Rules can be reflexive, allowing the construction of powerful meta-rules. HPRL runs on a Vax 11/780, and on the HP-9836. It has been used for various exploratory projects at Hewlett-Packard, including a program to diagnose faults during IC manufacturing, a program for analyzing dual-channel ECG information to diagnose arrhythmias, and a program for analyzing spectra from infrared and mass spectrometers.

Proceedings ArticleDOI
01 Jan 1983
TL;DR: In the last few years, expert systems have become the most visible and fastest growing branch of Artificial Intelligence and it is necessary to address issues of knowledge acquisition, knowledge representation, inference mechanisms, control strategies, user interface and dealing with uncertainty.
Abstract: In the last few years, expert systems have become the most visible and fastest growing branch of Artificial Intelligence. Their objective is to capture the knowledge of an expert in a particular problem domain, represent it in a modular, expandable structure, and transfer it to other users in the same problem domain. To accomplish this goal, it is necessary to address issues of knowledge acquisition, knowledge representation, inference mechanisms, control strategies, user interface and dealing with uncertainty.

Proceedings Article
22 Aug 1983
TL;DR: This research describes learning in intermediate domains, where causal knowledge is used in conjunction with experience to build new hypotheses and guide behavior, and causal knowledge of the domain is essential in order to create a correct explanation of a failure.
Abstract: Learning from prediction failures is one of the most important types of human learning from experience. In particular, prediction failures provide a constant source of learning. When people expect some event to take place in a certain way and it does not, they generate an explanation of why the unexpected event occurred [Sussman 1975] [Schank 1982]. This explanation requires hypotheses based on the features of the objects and on causal relations between the events in the domain. In some domains, causal knowledge plays a large role; in some, experience determines behavior almost entirely. This research describes learning in intermediate domains, where causal knowledge is used in conjunction with experience to build new hypotheses and guide behavior. In many cases, causal knowledge of the domain is essential in order to create a correct explanation of a failure. The HANDICAPPER program uses domain knowledge to aid it in building hypotheses about why thoroughbred horses win races. As the program processes more races, it builds and modifies its rules, and steadily improves in its ability to pick winning horses.


01 Sep 1983
TL;DR: This paper proposes a representation and inference rules for reasoning about belief and knowledge and describes the processes that create, store and use beliefs and knowledge in the formal system.
Abstract: : Artificial Intelligence programs must have common-sense knowledge. This includes knowledge about beliefs and knowledge. If we have a knowledge representation that can represent facts about beliefs and knowledge, and an adequate set of inference rules, we have taken the first step in building a program that can reason about beliefs and knowledge. The next step is to devise a search strategy: an algorithm that decides which inference rules to apply to which expressions to solve a problem. This paper proposes a representation and inference rules for reasoning about belief and knowledge. The core of the paper is a series of examples of representation and inference in the formal system. These examples describe the processes that create, store and use beliefs and knowledge. Perception, introspection, memory, inference and planning are all considered. Finally there is an appendix with proofs that the formalism works as claimed.

Book ChapterDOI
Werner Horn1
19 Sep 1983
TL;DR: ESDAT is a decision support system designed for the application in primary medical care that represents domain knowledge in a semantic net and utilizes detailed knowledge from the subfield rheumatology.
Abstract: ESDAT is a decision support system designed for the application in primary medical care. It represents domain knowledge in a semantic net. Instantiation of the concepts of the net is performed by activating nodes. They form hypotheses on a global blackboard. The inference process utilizes the hypothesize-and-test paradigm. Major emphasis is put on focussing mechanisms, which restrict the number of hypotheses the system is currently concentrating on. The current version of the system utilizes detailed knowledge from the subfield rheumatology.



01 Jan 1983
TL;DR: Getree-a new interactive knowledge management system-is being designed and implemented in GE Corporate Research, where the user-organization can have full responsibility for and control of the knowledge base.
Abstract: In practical expert systems, regardless of inference technology, the accuracy and completeness of the knowledge base determine expert system performance, and the cost of acquiring that knowledge base tends to dominate all other hardware and software costs. To reduce knowledge acquisition cost and error rate, Getree-a new interactive knowledge management system-is being designed and implemented in GE Corporate Research. In the new system, the knowledge base is represented as an (almost tree-like) network of nodes which can be displayed and manipulated on a personal computer workstation. Users-even unsophisticated users-can build and navigate these networks, modify nodes and connecting arcs, verify correctness visually, follow the execution of inference engines, and generate equivalent code to be run in more constrained target environments. Consequently, with Getree the user-organization can have full responsibility for and control of the knowledge base. 7 references.

Journal ArticleDOI
TL;DR: Aristotle's argument for the existence of non-demonstrative knowledge does not require the specification of, and hence the satisfaction of, any evidence condition as mentioned in this paper, and this is the case for all knowledge, whether basic or non-basic.
Abstract: As an introduction to explicating the concept of basic knowledge, I shall examine Aristotle's argument for the existence of basic knowledge and urge two basic points. The first point is that Aristotle's argument, properly viewed, establishes the existence of a kind of knowledge, basic or non-demonstrative knowledge, the definition of which does not require the specification of, and hence the satisfaction of, any evidence condition. This point has been urged by philosophers like Peirce and Austin but it needs further argumentation because most analytic epistemologists still insist (for reasons that we shall see) that all knowledge, whether basic or non-basic, requires the satisfaction of some evidence condition. Secondly, to urge (as Wittgenstein and Dewey have done) that the basic propositions whose existence is established by Aristotle's argument could be privileged but not known, for the reason that there is no evidence condition for them, would be to adopt a position that either entails wholesale skepticism or undermines the basic distinction between knowledge and belief.


Journal ArticleDOI
TL;DR: This paper proposes an approach to knowledge acquisition which is based on a deep restructuring of the usual architecture of an expert system, including the partitioning of knowledge in a rule base and a strategy-rule base, the treatment of partial matching, and the design of a specific knowledge acquisition subsystem.

ReportDOI
01 Dec 1983
TL;DR: This concept is developed and discussed in the context of proposed Navy training systems for acquiring models of trainee performance during learning, rules of behavior for an automated opponent in a tactics trainer, and a knowledge base of facts to be subsequently presented to trainees for memorization.
Abstract: : This reserach developed a design concept for an interactive system to acquire domain knowledge form a training expert. Such a system would facilitate the development of knowledge-based instructional systems by directly eliciting and encoding from domain experts knowledge needed to deliver instruction. An analysis of the process by which knowledge-based systems are constructed indicates that the generality of a knowlege acquisition system must be limited by domain characteristics and by the architecture of the system it serves, and the non-sequential, interacting activities during system development constrain the potential role of automated knowledge acquisition aids. A feasible concept for knowledge acquistion technology, building on prior research in artificial intelligence, involves the notion of class-generic systems for a related set of domains with a fixed architecture and training capabilities. This concept is developed and discussed in the context of proposed Navy training systems for acquiring models of trainee performance during learning, rules of behavior for an automated opponent in a tactics trainer, and a knowledge base of facts to be subsequently presented to trainees for memorization.

01 Dec 1983
TL;DR: The productive procedures used by experts were translated into specific guidelines toward improving circuit troubleshooting, and the effectiveness of these guidelines will be evaluated in a subsequent experiment to investigate individual differences in procedural knowledge.
Abstract: : This work is concerned with the knowledge that electronics technicians possess of electronic equipment, and more generally, with how people operate in tasks that draw upon a complex spatial symbolic knowledge base. A technician's knowledge base is postulated to consist of three types of related knowledge: structural/functional knowledge, which pertains to the actual configuration of a circuit and the role that its components play in the operation of the device; prototypical knowledge, which pertains to the general properties common to circuits of a given type; and procedural knowledge, which pertains to the way that a circuit can be modified and to the interaction among knowledge elements of all three types of knowledge. The present report focuses on an experiment conducted to investigate individual differences in procedural knowledge. Novice and expert subjects performed tasks in which they had to either locate and correct an error in a circuit, change the function of a circuit, or complete a missing segment in a circuit. On all tasks, experts were found to be far more accurate than novices; but more important, experts were classified -- on the basis of verbal protocols -- to be considerably more systematic, orderly and directed in their problem solving strategies. The productive procedures used by experts were then translated into specific guidelines toward improving circuit troubleshooting, and the effectiveness of these guidelines will be evaluated in a subsequent experiment. The results of this research program should help in providing guidelines for training electronic technicians to better understand and troubleshoot complex equipment.

Book ChapterDOI
01 Jan 1983
TL;DR: To construct a system with knowledge requires the ability to elicit knowledge and to construct an appropriate abstraction at a level of resolution adequate to deal with the set of expected situations for which the system was designed.
Abstract: Knowledge is examined from both the abstract and technological aspects. An approach to the abstract problem of defining knowledge is developed by distinguishing knowledge structures from meaning structures, and connecting knowledge with justification. The essentials of the three major knowledge representation schemes - semantic nets, production systems and frames - are examined. The conclusion is drawn that research is needed on the role of the users within a complete system. To construct a system with knowledge requires the ability to elicit knowledge and to construct an appropriate abstraction at a level of resolution adequate to deal with the set of expected situations for which the system was designed. New skills will have to be evolved and a new breed of computer expert (‘knowledge engineer’) will be required.

Book ChapterDOI
01 Jan 1983
TL;DR: Task-dependent knowledge contains syntax, semantics, and pragmatics and may be used to transform the surface structure of a recognized sentence into its conceptual representation.
Abstract: Task-dependent knowledge contains syntax, semantics, and pragmatics. Its purpose is the solution of ambiguities and recovery from errors which still remain at the lexical level. Furthermore, task-dependent knowledge may be used to transform the surface structure of a recognized sentence into its conceptual representation.

01 Jan 1983
TL;DR: A body of abstract control knowledge provides a generic framework for constructing knowledge bases for rclatcd problems in other domains and also provides a useful starting point for studying the nature of strategies.
Abstract: A poorly dcsigncd knowledge base can be as cryptic as an arbitrary program and just as difficult to maintain. Rcprescnting control knowledge abstractly, scparatcly from domain facts and relations, makes the design more transparent and explainable. A body of abstract control knowledge provides a generic framework for constructing knowledge bases for rclatcd problems in other domains and also provides a useful starting point for studying the nature of strategies.*