Showing papers on "Domain knowledge published in 1987"

I 3 R: a new approach to the design of document retrieval systems

Abstract: THE MOST EFFECTIVE METHOD OF IMPROVING THE RETRIEVAL PERFORMANCE OF A DOCUMENT RETRIEVAL SYSTEM IS TO ACQUIRE A DETAILED SPECIFICATION OF THE USER''S INFORMATION NEED. THE SYSTEM DESCRIBED IN THIS PAPER, (I(EXPONENT 3)R), PROVIDES A NUMBER OF FACILITIES AND SEARCH STRATEGIES BASED ON THIS APPROACH. THE SYSTEM USES A NOVEL ARCHITECTURE TO ALLOW MORE THAN ONE SYSTEM FACILITY TO BE USED AT A GIVEN STAGE OF A SEARCH SESSION. USERS INFLUENCE THE SYSTEM ACTIONS BY STATING GOALS THEY WISH TO ACHIEVE, BY EVALUATING SYSTEM OUTPUT, AND BY CHOOSING PARTICULAR FACILITIES DIRECT- LY. THE OTHER MAIN FEATURES OF (I(EXPONENT 3)R)) ARE AN EMPHASIS ON DOMAIN KNOWLEDGE USED FOR REFINING THE MODEL OF THE INFORMATION NEED, AND THE PROVISION OF A BROWSING MECHANISM THAT ALLOWS THE USER TO NAVIGATE THROUGH THE KNOWLEDGE BASE.

Knowledge base verification

Abstract: Publisher Summary This chapter describes several tools and techniques for validating a knowledge base, with emphasis on the rule checker. The other tools, such as the syntax checker and test case manager, also help the knowledge engineer to maintain the correctness of one's knowledge base and of the tool one is using. From the experiences with constructing different knowledge bases, one fined that many changes and additions are made during the development and maintenance of a knowledge base. The most frequent problems that CHECK has detected are unreachable and dead-end clauses. These types of errors are difficult to detect with conventional knowledge base debugging aids. It is found that redundant and conflicting rules appeared the least often. Nevertheless, CHECK'S automated facilities have saved engineers from many hours of tedious debugging. As the field of knowledge-based systems matures, large expert systems will be fielded in critical situations. As it will be impossible to test all paths beforehand, one must have assurance that deadly traps, such as circular rules and dead-end clauses do not exist in the knowledge base. Thus, verification facilities similar to the ones described in this paper becomes essential.

Knowledge representation for model management systems

Abstract: This paper examines the concept of a model management system, what its functions are, and how they are to be achieved in a decision support context. The central issue is model representation which involves knowledge representation and knowledge management within a database environment. The model abstraction structure is introduced as a vehicle for model representation which supports both heuristic and deterministic inferencing as well as the conceptual/external schema notion familiar to database management. The model abstraction is seen as a special instance of the frame construct in artificial intelligence. Model management systems are characterized as frame-systems and a database implementation of this approach is described.

Extracting expertise from experts: Methods for knowledge acquisition

Abstract: Knowledge acquisition is the biggest bottleneck in the development of expert systems. Fortunately, the process of translating expert knowledge to a form suitable for expert system development can benefit from methods developed by cognitive science to reveal human knowledge structures. There are two classes of these investigative methods, direct and indirect. We provide reviews, criteria for use, and literature sources for all principal methods. Direct methods discussed are: interviews, questionnaires, observation of task performance, protocol analysis, interruption analysis, closed curves, and inferential flow analysis. Indirect methods include: multidimensional scaling, hierarchical clustering, general weighted networks, ordered trees, and repertory grid analysis.

Use of a domain model to drive an interactive knowledge-editing tool

Abstract: The manner in which a knowledge-acquisition tool displays the contents of a knowledge base affects the way users interact with the system. Previous tools have incorporated semantics that allow knowledge to be edited in terms of either the structural representation of the knowledge or the problem-solving method in which that knowledge is ultimately used. A more effective paradigm may be to use the semantics of the application domain itself to govern access to an expert system's knowledge base. This approach has been explored in a program called OPAL, which allows medical specialists working alone to enter and review cancer treatment plans for use by an expert system called ONCOCIN. Knowledge-acquisition tools based on strong domain models should be useful in application areas whose structure is well understood and for which there is a need for repetitive knowledge entry.

Knowledge‐based organizations

Abstract: Organizations will increasingly be regarded as joint human‐computer knowledge processing systems. This perspective has significant implications for the design, management, and success of an organization. A knowledge‐based organization is seen as a society of knowledge workers who are interconnected by a computerized infrastructure. Their work with various distinct types of knowledge is supported in a coordinated, cooperative fashion by the computerized infrastructure. This article identifies the various knowledge types and the building blocks for an infrastructure. It suggests ways in which the blocks can be combined to enhance worker productivity and ways in which the traditional functional areas of management can contribute to the realization of viable knowledge‐based organizations.

Relative Importance of Domain Knowledge and Overall Aptitude on Acquisition of Domain-Related Information

Abstract: Two experiments investigated whether domain-specific expertise could compensate for low overall aptitude on certain domain-related cognitive processing tasks. It was hypothesized that the performance of low-aptitude individuals on a task requiring them to acquire new information in a domain would be a function of domain expertise rather than overall aptitude level. In Experiment 1, two low-aptitude groups, one with high domain knowledge and one with low domain knowledge, were presented with a baseball passage. On both recall and recognition tests, performance was a function of level of expertise in the domain. In Experiment 2, both level of baseball knowledge and overall aptitude were varied in a factorial design. Again, performance was a function of baseball knowledge rather than aptitude level. Low-aptitude/high-knowledge participants recalled more information than high-aptitude/low-knowledge participants. In addition, the performance of the low-aptitude/high-knowledge group was similar to the high-apti...

The problem of implicit knowledge

Abstract: Recognising the existence of different forms of knowledge is a first step towards effective knowledge elicitation. This article takes a brief look at some of the different types of knowledge which human experts possess and then focusses on the problem of implicit knowledge. The fact that much of an expert's knowledge is implicit or tacit in nature is a major problem for those working in the area of knowledge elicitation. Despite this, the topic has attracted little discussion or research. The present article reviews some of the limited literature on the topic and attempts to settle some of the confusion over what implicit knowledge is, or might be. Relevant experiments from the psychological literature are discussed. The paper also looks at possible ways of assessing implicit knowledge and makes recommendations for future research in this area.

Knowledge based tutor

Abstract: A knowledge system has a consultation system and also encodes domain-dependent tutoring knowledge as a set of conditions for interrupting the operation of the consultation system in order to evaluate a subject system. During the evaluation, the subject system is probed for its understanding of the status of the consultation system, and its understanding is compared to the actual status to obtain a measure of the subject system's knowledge and performance relative to that of the consultation system. The direction of the probing and the source of information for instruction or diagnosis is based upon the condition causing the interruption of the consultation system. Preferably an authoring system scans the domain-dependent knowledge base of the consultation system and determines a set of possible interrupt conditions. From this set a user exercises judgment in selecting a subset of conditions that are appropriate for the subject domain and the needs of the student. The selected conditions and a selected test case dialog are stored in a case file, and a number of different case files may be stored in a case library. The authoring system preferably creates a file or index of tutorial knowledge which correlates the relevant domain knowledge with the interrupt conditions. The tutorial knowledge includes, for example, expressions for causing interrupts after their values are found, rules concluding the expressions, the values concluded by the rules, and the factors in the rules.

KRITON: a knowledge-acquisition tool for expert systems

Abstract: A hybrid system for automatic knowledge acquisition for expert systems is presented. The system integrates artificial intelligence and cognitive science methods to construct knowledge bases employing different knowledge representation formalisms. For the elicitation of human declarative knowledge, the tool contains automated interview methods. The acquisition of human procedural knowledge is achieved by protocol analysis techniques. Textbook knowledge is captured by incremental text analysis. The goal structure of the knowledge elicitation methods is an intermediate knowledge-representation language on which frame, rule and constraint generators operate to build up the final knowledge bases. The intermediate knowledge representation level regulates and restricts the employment of the knowledge elicitation methods. Incomplete knowledge is laid open by patterndirected invocation methods (the intermediate knowledge base watcher) triggering the elicitation methods to supplement the necessary knowledge.

Distributed knowledge management based on product state models — the case of decision support in health care administration

Abstract: Knowledge management has inspired a shift from a transaction to a distributed knowledge management (DKM) perspective on inter-organizational information processing. The DKM concept structures the knowledge creation, knowledge sharing, and knowledge exploitation in organizations according to a product state model (PSM) required for management of technological diversity. Each player in the network acquires specific knowledge from other players for decision support. This article shows the relevance of the DKM model in a case study of a distributed decision support system (DDSS) in health care administration in the US.

The knowledge frontier: essays in the representation of knowledge

Abstract: 1. What Is Knowledge Representation?.- 1.1. Introduction.- 1.2. Logic representations.- 1.2.1. Default logic.- 1.2.2. Fuzzy logic.- 1.3. Semantic networks.- 1.3.1. Partitioned networks.- 1.3.2. Marker propagation schemes.- 1.3.3. Topic hierarchies.- 1.3.4. Propositional networks.- 1.3.5. Semantic networks and logic.- 1.4. Procedural representations.- 1.4.1. Winograd's work.- 1.4.2. Procedural semantic networks.- 1.5. Logic programming.- 1.6. Frame-based representations.- 1.7. Production system architectures.- 1.8. Knowledge representation languages.- 1.8.1. KL-One.- 1.8.2. KRYPTON.- 1.8.3. Other languages.- 1.9. Concluding remarks.- 2. Knowledge Representation: What's Important About It?.- 2.1. Introduction.- 2.2. Knowledge for reasoning agents.- 2.3. Modeling the external world.- 2.4. Perception and reasoning by machine.- 2.5. The Nature of the world and its models.- 2.6. The functions of a knowledge representation system.- 2.7. The knowledge acquisition problem.- 2.8. The perception problem.- 2.9. Planning to act.- 2.10. Role of a conceptual taxonomy for an intelligent agent.- 2.11. The structure of concepts.- 2.12. An example of a conceptual taxonomy.- 2.13. The need for taxonomic organization.- 2.14. Recognizing/analyzing/parsing situations.- 2.15. Two aspects of knowledge representation.- 2.16. Expressive adequacy.- 2.17. Notational efficacy.- 2.18. The relationship to formal logic.- 2.19. Concepts are more than predicates.- 2.20. Conclusions.- 3. Some Remarks on the Place of Logic in Knowledge Representation.- 3.1. Introduction.- 3.2. What is logic?.- 3.3. On being logical.- 3.4. Reasoning and logic.- 3.5. Nonmonotonic logic.- 3.6. Conclusion.- 4. Logic and Natural Language.- 4.1. Introduction.- 4.2. Default logic for computing presuppositions.- 4.3. Modal logic for planning utterances.- 4.4. Temporal logic for reasoning about futures.- 4.5. Conclusion.- 5. Commonsense and Fuzzy Logic.- 5.1. Introduction.- 5.2. Meaning representation in test-score semantics.- 5.3. Testing and translation rules.- 5.3.1. Composition of elastic constraints.- 5.4. Representation of dispositions.- 5.5. Reasoning with dispositions.- 5.6. Concluding remark.- 6. Basic Properties of Knowledge Base Systems.- 6.1. Introduction.- 6.2. Basic notions.- 6.3. Completeness & consistency of rule-represented knowledge bases.- 6.4. The case of linear sets of rules.- 6.5. Dependency of rules on attributes.- 6.6. Partial information and defaults.- 6.7. Conclusion.- 7. First Order Logic and Knowledge Representation: Some Problems of Incomplete Systems.- 7.1. Introduction.- 7.2. Prolog & Absys: declarative knowledge manipulation systems.- 7.3. Primitive goal selection strategies in Absys and Prolog.- 7.4. Selection strategies and knowledge systems.- 7.5. Summary.- 8. Admissible State Semantics for Representational Systems.- 8.1. Introduction - the problem of practical semantics.- 8.2. Internal and external meanings.- 8.3. Admissible state semantics.- 8.4. Example: semantic networks.- 8.5. Example: k-lines.- 8.6. Conclusion.- 9. Accelerating Deductive Inference: Special Methods for Taxonomies, Colours and Times.- 9.1. Introduction.- 9.2. Recognizing type relationships.- 9.3. Recognizing part-of relationships.- 9.4. Recognizing colour relationships.- 9.5. Recognizing time relationships.- 9.6. Combining general and special methods.- 9.7. Concluding remarks.- 10. Knowledge Organization and Its Role in Temporal and Causal Signal Understanding: The ALVEN and CAA Projects.- 10.1. Introduction.- 10.2. The representational scheme.- 10.2.1. Knowledge packages: classes.- 10.2.2. Knowledge organization.- 10.2.3. Multi-dimensional levels of detail.- 10.2.4. Time.- 10.2.5. Exceptions and similarity relations.- 10.2.6. Partial results and levels of description.- 10.3. The interpretation control structure.- 10.4. The ALVEN project.- 10.4.1. Overview.- 10.4.2. LV dynamics knowledge and its representation.- 10.5. The CAA project.- 10.5.1. Overview.- 10.5.2. Representation of causal connections.- 10.5.3. Use of causal links.- 10.5.4. Recent research related to causality.- 10.5.5. Representation of domain knowledge.- 10.5.6. Knowledge-base stratification and projection links.- 10.5.7. Recognition strategies and control.- 10.6. Conclusions.- 11. SNePS Considered as a Fully Intensional Propositional Semantic Network.- 11.1. Introduction.- 11.1.1. The SNePS environment.- 11.1.2. SNePS as a knowledge representation system.- 11.1.3. Informal description of SNePS.- 11.2. Intensional knowledge representation.- 11.3. Description of SNePS/CASSIE.- 11.3.1. CASSIE - A model of a mind.- 11.3.2. A conversation with CASSIE.- 11.3.3. Syntax and semantics of SNePS/CASSIE.- 11.3.4. The conversation with CASSIE, revisited.- 11.4. Extensions and applications of SNePS.- 11.4.1. SNePS as a database management system.- 11.4.2. Address recognition for mail sorting.- 11.4.3. NEUREX.- 11.4.4. Representing visual knowledge.- 11.4.5. SNeBR: A belief revision package.- 11.5. Knowledge-based natural language understanding.- 11.5.1. Temporal structure of narrative.- 11.6. Conclusion: SNePS and SNePS/CASSIE as Semantic Networks.- 11.6.1. Criteria for semantic networks.- 11.6.2. SNePS and SNePS/CASSIE vs. KL-One.- 12. Representing Virtual Knowledge Through Logic Programming.- 12.1. Introduction.- 12.2. Representing knowledge in Prolog.- 12.3. Asking for inferences - virtual knowledge.- 12.4. Representing problem-solving knowledge.- 12.5. Representing database knowledge.- 12.6. Limitations.- 12.7. Conclusions.- 13. Theorist: A Logical Reasoning System for Defaults and Diagnosis.- 13.1. Introduction.- 13.2. Prolog as a representation system.- 13.3. The Theorist framework.- 13.4. Tasks appropriate for the Theorist framework.- 13.4.1. Nonmonotonic reasoning - reasoning with default and generalised knowledge.- 13.4.2. Diagnosis.- 13.4.3. Learning as theory construction.- 13.4.4. User modelling as theory maintenance.- 13.4.5. Choices in mundane tasks.- 13.5. Representation and reasoning in theorist.- 13.5.1. Extending Horn clauses to full first order logic.- 13.5.2. Reasoning as the construction of consistent theories.- 13.6. Implementing a Theorist prototype in Prolog.- 13.6.1. Not parallelism.- 13.7. Status and conclusions.- 14. Representing and Solving Temporal Planning Problems.- 14.1. Introduction.- 14.2. The Time Map Manager.- 14.2.1. A Predicate Calculus Database.- 14.2.2. Adding Basic Concepts of Time.- 14.2.3. Events and Persistences.- 14.2.4. Temporal Database Queries.- 14.2.5. Chaining Rules in a Temporal Database.- 14.2.6. A Simple Planner Based on the TMM.- 14.3. The Heuristic Task Scheduler.- 14.3.1. Describing a Resource.- 14.3.2. Describing a Plan.- 14.3.3. Specifying Plan Resource Use.- 14.3.4. Specifying Plan Tasks.- 14.3.5. Specifying Plan Constraints.- 14.3.6. Producing a Completed Linear Task Ordering.- 14.4 Summary and Conclusions.- 15. Analogical Modes of Reasoning and Process Modelling.- 15.1. Introduction to analogical reasoning.- 15.2. WHISPER: A program using analogs.- 15.3. Observations on the use of analogs.- 15.4. Mental rotation as an analog process.- 15.5. Conclusions.- 16. Representing and Using Knowledge of the Visual World.- 16.1. Introduction.- 16.2. Progress in high-level vision.- 16.3. The complexity barrier.- 16.4. Achieving descriptive adequacy.- 16.5. Achieving procedural adequacy.- 16.6. Conclusion.- 17. On Representational Aspects of VLSI-CADT Systems.- 17.1. Introduction.- 17.2. VLSI design process.- 17.2.1. Use of multiple perspectives.- 17.2.2. Almost hierarchical design.- 17.2.3. Constraints and partial specifications.- 17.3. VLSI design knowledge.- 17.3.1. Knowledge about VLSI design.- 17.4. VLSI design representation.- 17.4.1. Representation of designed artifact.- 17.4.2. Design plan.- 17.5. Analysis, testing, and diagnosis of VLSI circuits.- 17.5.1. Reasoning with constraints.- 17.5.2. Qualitative analysis.- 17.5.3. Design for testability frames.- 17.5.4. Logic programming in VLSI design.- 17.5.5. Diagnostic reasoning.- 17.6. Natural language interfaces.- 17.7. Concluding remarks.

Design for acquisition: principles of knowledge-system design to facilitate knowledge acquisition

Abstract: The problem of knowledge acquisition is viewed in terms of the incongruity between the representational formalisms provided by an implementation (e.g. production rules) and the formulation of problem-solving knowledge by experts. The thesis of this paper is that knowledge systems can be designed to facilitate knowledge acquisition by reducing representation mismatch. Principles of design for acquisition are presented and applied in the design of an architecture for a medical expert system called MUM. It is shown how the design of MUM makes it possible to acquire two kinds of knowledge that are traditionally difficult to acquire from experts: knowledge about evidential combination and knowledge about control. Practical implications for building knowledge-acquisition tools are discussed.

Artificial intelligence and software engineering

Abstract: Software Engineering is a knowledge-intensive activity, requiring extensive knowledge of the application domain and of the target software itself. Many Software Engineering costs can be attributed to the ineffectiveness of current techniques for managing this knowledge, and Artificial Intelligence techniques can help alleviate this situation. More than two decades of research have led to many significant theoretical results, but few demonstrations of practical utility. This is due in part to the amount and diversity of knowledge required by Software Engineering activities, and in part to the fact that much of the research has been narrowly focused, missing many issues that are of great practical importance. Important issues that remain to be addressed include the representation and use of domain knowledge and the representation of the design and implementation history of a software system. If solutions to these issues are found, and experiments in practical situations are successful, the implications for the practice of Software Engineering will be profound, and radically different software development paradigms will become possible.

Learning by Experimentation

Abstract: Autonomous systems require the ability to plan effective courses of action under potentially uncertain or unpredictable contingencies. Effective planning requires knowledge of the environment, and if the environment is too complex or changes dynamically, goal-driven learning with reactive feedback becomes a necessity. This paper addresses the issue of learning by experimentation as an integral component of PRODIGY, a flexible planning system augmented with capabilities for execution monitoring and dynamic replanning upon receiving adverse feedback. In particular, experiment formulation seeks to acquire precisely the domain knowledge needed to complete a partial plan, or to correct an errant one. Thus, experimentation is demand-driven and exploits both the internal state of the planner and any external feedback received. A detailed example of integrated experiment formulation in presented as the basis for a systematic approach to extending an incomplete domain theory or correcting a potentially inaccurate one. 1

The Knowledge Acquisition Grid: a method for training knowledge engineers

Abstract: This paper describes the Knowledge Acquisition Grid, developed to assist knowledge engineers in the manual transfer of expertise. The Grid is used in a knowledge-acquisition module which itself is part of a larger program designed to train people in knowledge engineering techniques offered by Digital Equipment Corporation. The Grid describes a two-dimensional space in which five forms of expert knowledge and six basic types of interview questions constitute the horizontal and vertical dimensions respectively. Description of the rationale, dimensions, components, and strategy for use of the Grid in the knowledge-acquisition component of building an expert system is provided along with discussion of the need for greater attention in general to the social psychology of expert interviewing.

Knowledge systems and PROLOG

Abstract: Every word to utter from the writer involves the element of this life. The writer really shows how the simple words can maximize how the impression of this book is uttered directly for the readers. Even you have known about the content of knowledge systems and prolog so much, you can easily do it for your better connection. In delivering the presence of the book concept, you can find out the boo site here.

Conceptual organization of case law knowledge bases

Abstract: Conceptual retrieval requires the computer to have knowledge of legal concepts and issues, and their relationship to the case law collection. This paper discusses the organization of a case law knowledge base in terms of three interacting components: a domain knowledge model defines the basic concepts of a case law domain; individual case descriptors describe the particular constellation of concepts that pertain to each case, organized into a frame-based superstructure according to the legal roles they fill; and issue/case discrimination trees represent the significance of each case relative to a model of the normative relationships of the legal domain. Each of these components is described and justified by showing its contribution to the goal of conceptual retrieval.

KNACK—report-driven knowledge acquisition

Abstract: This paper describes a knowledge-acquisition tool that builds expert systems for evaluating designs of electro-mechanical systems. The tool elicits from experts (1) knowledge in the form of a skeletal report, (2) knowledge about a large collection of report fragments, only some of which will be relevant to any specific report, and (3) knowledge of how to customize the report fragments for a particular application. The tool derives its power from exploiting its understanding of two problem-solving methods and of the different roles that knowledge plays in those two methods.†

Different Techniques and Different Aspects on Declarative Knowledge

Abstract: In this chapter we describe techniques for discovering organization in an expert’s domain knowledge. These techniques are illustrated in the domain of “domestic gas-fired hot water and central heating systems,” which possesses technical properties seen as relevant to larger-scale domains. The informant was not a recognized expert on central heating but a scientist with an interest in the domain.

A logical approach to reasoning by analogy

Abstract: We analyze the logical form of the domain knowledge that grounds analogical inferences and generalizations from a single instance. The form of the assumptions which justify analogies is given schematically as the "determination rule", so called because it expresses the relation of one set of variables determining the values of another set. The determination relation is a logical generalization of the different types of dependency relations denned in database theory. Specifically, we define determination as a relation between schemata of first order logic that have two kinds of free variables: (1) object variables and (2) what we call "polar" variables, which hold the place of truth values. Determination rules facilitate sound rule inference and valid conclusions projected by analogy from single instances, without implying what the conclusion should be prior to an inspection of the instance. They also provide a way to specify what information is sufficiently relevant to decide a question, prior to knowledge of the answer to the question.

Validation of knowledge-based systems

Abstract: Introduction Knowledge-based systems (KBSs) have become ever more important in the development of large, complex software systems in defense, industry, business, and science. Failures in these systems may gravely endanger human life and property. "Validation" of KBSs thus greatly concerns software systems developers and implementors. This still relatively young research area' requires a methodology distinct from traditional software verification, validation, and testing (VV&T) technology, for two reasons: the nature of the problems to be solved and the lack of explicit functional structure in KBS. In traditional software systems development the problem to be solved is typically well-understood and the requirements statements for the problem exist and are precise and well-defined. KBSs are typically used when the problem is not well-understood, and thus lacks precise requirements statements. Moreover, traditional software is procedural, with explicit control flow within and between modules. KBS software, in contrast, is mostly declarative and nondeterministic, with little or no explicit control and no explicit functional modules. It lacks the explicit functional structure prerequisite to requirements validation. KBSs overcome the first problem through the so-called incremental development paradigm, also referred to as "exploratory programming" and "rapid prototyping." The system designer translates the imprecise, incomplete, and often incorrect requirements into a partial, incomplete but executable KBS written in an expert system (ES) shell, such as ART, OPS5, or KEE. (See the chapter by W. B. Gevarter elsewhere in this volume.) Based on the output, the KBS is modified (corrected, deleted or extended) until the

Knowledge-based software design using design schemas

Abstract: Design schemas provide a means for abstracting software designs into broadly reusable components that can be assembled and refined into new software designs. This paper describes a knowledge-based software development paradigm that is based on the design schema representation. It combines design schemas, domain knowledge, and various types of rules to assist in the quick generation of software designs from user specifications. A prototypical environment, IDeA (Intelligent Design Aid), is described that supports the knowledge-based paradigm. The schema-based techniques used in IDeA are presented along with some examples of their use.

Cooperating knowledge-based assistants for the office

Abstract: This paper presents an approach to high-level support of office workers by embedding office knowledge in a network of distributed cooperating knowledge-based or expert “assistants” and servers. These knowledge-based systems incorporate both factual and procedural knowledge and are capable of making use of existing conventional office technology. They constitute a form of computer-supported cooperative work. We describe a common architecture for our assistants and servers that incorporates several key features. Our systems are capable of supporting concurrent multiple consultations or tasks and have facilities for the interruption and resumption of consultations as appropriate. The various assistants and servers, which may reside on different machines, cooperate in solving problems or completing tasks by passing messages. We propose a taxonomy of the general office knowledge normally used by office workers, together with a frame and rule-based knowledge representation scheme. We also describe an experimental system, written in PROLOG, that incorporates the above design principles.

Knowledge-based knowledge acquisition for a statistical consulting system

Abstract: Knowledge-based knowledge acquisition means restricting the domain of knowledge that can be acquired and developing a conceptual model of the domain. We have built a prototype knowledge-based knowledge acquisition system for the domain of data analysis. A critique of the prototype has led to a design for a possibly practical data analysis knowledge acquisition system.

A machine learning approach to the automatic synthesis of mechanistic knowledge for engineering decision-making

Abstract: Inductive learning is proposed as a tool for synthesizing domain knowledge from data generated by a model-based simulator. In order to use an inductive engine to generate decision rules, the pre-classification process becomes more complicated in the presence of multiple competing objectives. Instead of relying on a domain expert to perform this pre-classification task, a clustering algorithm is used to eliminate human biases involved in the selection of a classification function for pre-classification. It is shown that the use of a clustering algorithm for pre-classification not only further automates the process of knowledge by synthesizing, but also improves the quality of the rules generated by the inductive engine.