By Topic

AI Systems in Government Conference, 1989.,Proceedings of the Annual

Date 27-31 March 1989

Filter Results

Displaying Results 1 - 25 of 47
  • Proceedings of the Annual AI Systems in Government Conference (IEEE Cat. No.89CH2715-1)

    Publication Year: 1989
    Save to Project icon | Request Permissions | PDF file iconPDF (115 KB)  
    Freely Available from IEEE
  • Human corps planning: a first-order computational model

    Publication Year: 1989 , Page(s): 38 - 43
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (444 KB)  

    Human authors have attempted to identify a plausible first-order model which characterizes the high-level control cycle of expert human planners in the task of battlefield planning. Literature on the training of battlefield planning experts, interviews with expert planners, Army doctrine, and observations of planning activities of human experts were used to construct the model. The model is evaluated against a verbal protocol of a simulated corps maneuver-control planning exercise conducted at the US Army War College. It is concluded that the most striking aspect of the protocol for this planning problem is the interleaving of the task of subproblem formulation with the tasks of subproblem decomposition, plan critiquing, and plan repair. The evidence from the protocol is shown to be consistent with the ordering of these tasks as predicted from the first-order model View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Diagnosing an analog feedback system using model-based reasoning

    Publication Year: 1989 , Page(s): 290 - 295
    Cited by:  Papers (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (468 KB)  

    The approach developed by J. de Kleer and B.C. Williams (1987) has been extended to diagnose steady-state faults in an analog feedback system. A prototype system called AMFI (automated model-based fault isolation) is described which automatically diagnoses a target system given only information on the connectivity and component transfer functions. The prototype detects inconsistencies among measurements, calculates single- and multiple-fault hypotheses and their probabilities, and suggests the best next measurement View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Policy analysis and decision aids for energy crisis management

    Publication Year: 1989 , Page(s): 44 - 49
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (332 KB)  

    The authors describe the feature of a knowledge-based system which has been prototyped and demonstrated for the USA Office of Energy Emergencies but not yet carried through to the status of a production system. A decision aid has been prototyped in ARITY/Prolog which efficiently guides the microcomputer user through the development of an energy crisis scenario, the available policy options, the action choices, and the legal justifications for such actions. Future directions and motivations for those directions are also discussed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An automated injury coding system

    Publication Year: 1989 , Page(s): 24 - 28
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (344 KB)  

    The authors describe a prototype expert system for assigning codes to injuries detected in medical records such as hospital discharge summaries and coroner's reports. The system is intended to assist investigators in the task of analyzing official data related to injuries sustained by the occupants of automobiles involved in highway accidents. The design draws on previous work by the authors in the areas of knowledge-based information retrieval and pragmatically oriented natural-language processing. The prototype system is implemented on a high-performance workstation and can correctly identify and code a range of injury types View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Environs-an expert system: application to the impact of environmental pollution on the health of industrial workers and general public

    Publication Year: 1989 , Page(s): 33 - 36
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (232 KB)  

    An interactive system called Environs which enables experts gradually to construct a knowledge base is described. This rule-based expert system makes it possible for an expert to convey his expertise to the system without the intervention of a computer specialist. This knowledge-based production rule system is constructed to reduce the superfluity of rules and eliminate inconsistency, thereby securing maximum compatibility of the rules entered by the expert. The expert system is rule-interpreted with an optimized dialogue in pseudonatural language using the already-known facts. The system is tested on the impact of environmental pollution on the health of industrial workers and the general public. The application part is designed to detect a pollutant and its source or origin, and to help choose necessary steps in environmental pollution management View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Large knowledge based systems: an efficient approach

    Publication Year: 1989 , Page(s): 58 - 64
    Cited by:  Papers (1)  |  Patents (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (516 KB)  

    The authors propose an architecture for an integrated data/knowledge-based system (KBS) centered on a global data/knowledge dictionary and directory. The process of developing a KBS based on Prolog implementation schemes was studied an an architecture proposed which allowed the extensional (fact) portion of the knowledge base to exceed the KBS-resident system's memory. Under this architecture approach, the extensional portion of the knowledge base was integrated with the databases into a combined data/knowledge based system View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The DBG message understanding system

    Publication Year: 1989 , Page(s): 258 - 265
    Cited by:  Papers (1)  |  Patents (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (540 KB)  

    The DBG (Data Base Generator) message understanding system takes as input the text of military messages of various types and produces as output a representation of the content of the messages in the form of instantiated templates, which are frame structures specifying the properties of entities and events in particular domains. These structures can then be used to generate sets of data elements for specific downstream applications, such as automated knowledge base/database update from text. The authors summarize current capabilities and distinctive features of the system and describe the results of a test involving the generation of data vectors for a space situation assessment expert system from the text of space event messages View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Reactive planning using a `situation space'

    Publication Year: 1989 , Page(s): 50 - 55
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (512 KB)  

    The problem of how to plan in tactical situations (such as anti-submarine warfare) where planning must be responsible to events, or other agent's actions, which lie outside the predictive capability of the planner is addressed. The basic difficulty presented by this type of planning problem is how to engage in actions that are coherently related to the achievement of the overall goal despite the fact that often the planner cannot develop a complete plan for achieving that overall goal. To overcome this difficulty, the authors have developed a planning model within which the planner is controlled by knowledge organized into what is termed a situation space. The situation space guides the selection of goals and the construction of complete subplans which are appropriate to situations that arise and are coherently related to the overall goal. The situation space supports the principled generation of plan failure and replanning in a reactive environment View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Expert systems as job training aids

    Publication Year: 1989 , Page(s): 113 - 119
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (528 KB)  

    The author presents a proposal for the development of a general training aid that can be used with expert job aids to enhance on-the-job training. The training aid is based on a goal-directed architecture and can interact in several ways with a trainee, either passively, actively, or responsively. The focus for an initial test of the training aid will be procedure training based on an expert procedural reasoning system. It is concluded that the goal-directed design will make it possible to explore various ways of determining relevance and presenting information. For example, there is currently no connection between the procedural knowledge and a schematic representation of the system. Plans that achieve the goal of justifying a step in a procedure could be modified to utilize a schematic for explanation when one is available View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Compiling phrase structure grammar rules into register vector grammar

    Publication Year: 1989 , Page(s): 244 - 249
    Cited by:  Patents (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (392 KB)  

    A procedure is described for compiling context-free phrase structure grammar (PSG) rules into equivalent context-free register vector grammars (RVG). The procedure makes use of finite-state automata (FSA) as an intermediate form, as well as standard FSA transformation and reduction algorithms to minimize the size of the resulting register vector grammars. An example using the syntax of English auxiliary verbs demonstrated the use of FSA in the translation process, and it was shown that such automata must be reduced to yield fewer RVG productions. It was also shown that an additional optimization phase is needed to further reduce the size and increase the efficiency of register vector grammars. The PSG-to-RVG compiler enables natural-language interface developers to design grammars at the phrase structure rule level, but execute them at the register vector grammar level. This approach maximizes both the understandability and efficiency of syntactic analysis View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A hybrid approach to target tracking

    Publication Year: 1989 , Page(s): 80 - 87
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (540 KB)  

    The author describes a proposed AI (artificial intelligence) system for tracking the activity of a known set of targets given a time sequence of reports which contain clues to targets' identity and position. The system is oriented toward tracking multiple highly agile targets; continuous motion is not assumed. An example concerns bird tracking. The system is hybrid in that it combines a production system with a Bayesian belief network. The author outlines the system architecture, along with some of the motivations which led to it View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • PX1: a space shuttle mission operations knowledge-based system project

    Publication Year: 1989 , Page(s): 175 - 178
    Cited by:  Patents (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (316 KB)  

    A knowledge-based system (KBS) prototyping project is examined. The work entails reasoning about the impact of component failures in space shuttle orbiter subsystems. The effort is directed toward the development of a system that will recognize passive component failures as potential safety hazards, rather than toward an active failure identification or diagnostic tool. The system was designed to be integrated as a knowledge-based processing system utilizing input from a procedure-based processing system. Implementation of the consolidated system will occur when real-time telemetry data is available at the workstations. The system may be used standalone in the meantime. Flight controllers at the Johnson Space Center will use this prototype to help develop requirements for a space shuttle mission operations tool View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Explanation-based learning of diagnostic heuristics: a comparison of learning from success and failure

    Publication Year: 1989 , Page(s): 164 - 168
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (444 KB)  

    The author compares strategies of learning from failures to learning from successes in the context of a generate-and-test problem solver. One result is fairly straightforward: failure-driven learning creates rules which distinguish between failures. This is demonstrated by the fact that the number of hypotheses decreases after learning. A more subtle result is that the performance of the system, measured in terms of logical inferences, decreased with failure-driven learning more than it did with two variants of success driven learning. Diagnosis results are presented for ACES designed to process telemetry data from a satellite and isolate the cause of problems with the attitude control system View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Probable implication, the contrapositive, and contingency tables

    Publication Year: 1989 , Page(s): 9 - 13
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (420 KB)  

    In everyday life, the deductive logic of mathematical reasoning is generally replaced by inductive logic, which usually derives uncertain conclusions from a body of observed facts. The authors examine this distinction using probabilities based on 2×2 contingency tables. Some anomalies are discussed, such as the difference between the empirical probability of the implication A=>B and that of the logically equivalent contrapositive implication not-B=> not-A, although they are almost equal under near-certainty. Other classical examples are also discussed quantitatively, such as modus ponens and transitivity of implication View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • AI in the IRS

    Publication Year: 1989 , Page(s): 226 - 232
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (576 KB)  

    The IRS (US Internal Revenue Service) entered the field of AI (artificial intelligence) by training 13 domain experts; on completion of their training, these employees returned to the IRS AI Lab, where they identified, developed, and managed AI projects. The history of the IRS AI program is reviewed. In addition to applications currently under development, potential future projects are also identified View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Automated knowledge generation in support of Shuttle ground operations

    Publication Year: 1989 , Page(s): 268 - 274
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (480 KB)  

    A system is described that performs automated knowledge generation (AKG) using structural databases to create semantic knowledge bases. The approach, borrowed from machine vision, uses relaxation labeling to propagate constraints through the system. A prototype system has been developed which accesses an Intergraph CAD database, `processes' the information contained therein, and generates a frame-based knowledge base usable by NASA's Knowledge-based Autonomous Test Engineer (KATE) diagnostic system. The authors present a summary of work to date and discuss some of the obstacles found by the research group during the development of the AKG system View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Learning from historical precedent

    Publication Year: 1989 , Page(s): 150 - 156
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (520 KB)  

    Explanation-based learning, a method of abstracting general principles from a small number of prior cases, is discussed. The author demonstrates the feasibility of applying this method to economic sanction incidents. This approach is contrasted with regression analysis, a traditional quantitative method. A method for integrating these two approaches is proposed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Applying knowledge-based expert system to meat grading

    Publication Year: 1989 , Page(s): 120 - 123
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (324 KB)  

    The author presents an ongoing project, which has the immediate objective of developing a knowledge-based expert system to assist meat graders and an ultimate objective of automating the meat grading process through applications of image processing, pattern recognition, and expert systems. A voice-input, knowledge-based system for grading carcass beef has been developed and tested. The meat grader provides input for the system in the form of carcass characteristics by speaking into a microphone headset. on reception of the characteristics of the carcass, the knowledge-based system reasons through the production rules to reach quality and yield grades for the carcass View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An assistant expert system: assisting assistors in assisting taxpayers

    Publication Year: 1989 , Page(s): 210 - 217
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (660 KB)  

    The US Internal Revenue Service (IRS) employs 5000 assistors to answer telephone inquiries from taxpayers. The author describes the design and development of an expert system, the Taxpayer Service Assistant (TSA), that helps assistors provide advice to taxpayers on tax law topics. As measured by the GAO (US Government Accounting Office), these assistors provide correct and complete answers on only 63% of the 18 million calls on tax law that are received yearly. The IRS is exploring text retrieval and expert systems as alternative methodologies for supporting assistors and improving the quality of service. The author examines the tradeoffs made by each approach and shows why expert systems are more likely to be effective View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Knowledge acquisition for an Internal Revenue Service classification system

    Publication Year: 1989 , Page(s): 281 - 288
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (636 KB)  

    The authors introduce the Automated Issue Identification System (AIIS), an expert system prototype designed and built for the US Internal Revenue Service (IRS). They describe the classification task as currently performed by IRS auditors, and the expected benefits of automating this task with an expert system. The multilayer objectives of the knowledge engineering team are specifically detailed. Knowledge engineering techniques and procedures used in the initial phases of the project are examined, including the advantages and challenges of working with several experts. The main phases of the knowledge engineering process are described, from the initial task analysis, through early modeling of the IRS experts by the system, to the very effective rapid prototyping cycle supported by the object-oriented programming environment used in this project. The authors concludes with the current status of the prototype and the knowledge acquisition process, and a discussion of projected goals View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The PUNDIT natural-language processing system

    Publication Year: 1989 , Page(s): 234 - 243
    Cited by:  Papers (1)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (644 KB)  

    The authors describe the PUNDIT (Prolog Understanding of Integrated Text) text-understanding system, which is designed to analyze and construct representations of paragraph-length text. PUNDIT is implemented in Quintus Prolog, and consists of distinct lexical, syntactic, semantic, and pragmatic components. Each component draws on one or more sets of data, including a lexicon, a broad-coverage grammar of English, semantic verb decompositions, rules mapping between syntactic and semantic constituents, and a domain model. Modularity, careful separation of declarative and procedural information, and separation of domain-specific and domain-independent information all contribute to a system which is flexible, extensible and portable. Versions of PUNDIT are now running in five domains, including four military and one medical View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Prospector II

    Publication Year: 1989 , Page(s): 88 - 92
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (360 KB)  

    Prospector II has been designed as a frame-based expert system and the successor to PROSPECTOR to assist the geologist in performing regional mineral resource assessments. The changes that have occurred relate to the role of descriptive mineral deposit models in resource assessment and to the arrival of graphics-oriented workstations. These changes have brought about an expansion of the volunteer mode, the glossary of geological terms, and the explanation facilities in PROSPECTOR. The information volunteered by the geologist during a session with Prospector II is used to identify the deposit model that best fits the observations and to provide an explanation of what is explained or not explained by the model, and to indicate what is missing in the observations with respect to the model. The advice that is offered gives the geologist greater confidence in choosing a model and in deciding what additional data are needed to perform an assessment in an area View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Knowledge-based intrusion detection

    Publication Year: 1989 , Page(s): 102 - 107
    Cited by:  Papers (11)  |  Patents (63)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (504 KB)  

    The authors describe the expert-system aspects of IDES (intrusion-detection expert system). A system for computer intrusion detection IDES uses two distinct approaches to detect anomalies (which could signify intrusions) in a computer system, namely, statistical and rule-based anomaly detection. In the statistical approach, recent behavior of a subject of a computer system is compared with observed behavior and any significant deviation is considered anomalous. In the rule-based approach, acceptable behaviour of a subject is captured by a set of rules which is used to identify anomalous observed behavior. The authors claim that integrating the two approaches in IDES provides for a comprehensive system for detecting intrusions as they occur View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • OPERA-an expert operations analyst for a distributed computer system

    Publication Year: 1989 , Page(s): 179 - 185
    Cited by:  Papers (1)  |  Patents (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (612 KB)  

    OPERA (operations analyst) is a suite of expert systems being developed to improve operations support of a complex distributed computer network in the Space Shuttle Launch Processing System. The two expert systems in the initial OPERA prototype assist users in isolating and correcting network problems. One expert system monitors network error messages, which interprets to hypothesize faults and to suggest troubleshooting and problem recovery procedures. The second system searches a knowledge base containing symbolic digests of error tracking databases, retrieving relevant precedents to network anomalies (detected by the first system). Expert systems are integrated into OPERA by embedding them in blackboard structures. A controller blackboard routes communications among expert systems and coordinates their activities. The architecture supports shared access to external interfaces as well as domain knowledge bases, and is highly extensible View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.