The author explores a general framework for analyzing or synthesizing intelligent systems, as well as techniques for implementing them. The framework consists of six factors: purpose, space, structure, time, process a...
详细信息
ISBN:
(纸本)0818620595
The author explores a general framework for analyzing or synthesizing intelligent systems, as well as techniques for implementing them. The framework consists of six factors: purpose, space, structure, time, process and efficiency. The technologies involve the methods of artificial intelligence and strategies for developing knowledge-based systems. The framework has been used to develop diverse applications, ranging from an expertsystem for creative product design to a package for factory configurations. This framework is discussed in the context of an expertsystem for the design of personal computers.
The author discusses the experiences at the Idaho National Engineering Laboratory in developing an expertsystems/advanced computer technology organization. A brief history of the group is presented with discussions o...
详细信息
ISBN:
(纸本)0818620595
The author discusses the experiences at the Idaho National Engineering Laboratory in developing an expertsystems/advanced computer technology organization. A brief history of the group is presented with discussions of personnel, training, hardware, and software. Two example applications from the laboratory are briefly described to point out some lessons learned. The author presents several opinions/lessons learned in developing expertsystems capability and applying expertsystems technology.
A strategy is presented for transferring expertsystem technology from university to industry. This strategy is based on five years of experience in developing expert and knowledge-based systems for academic and indus...
详细信息
ISBN:
(纸本)0818620595
A strategy is presented for transferring expertsystem technology from university to industry. This strategy is based on five years of experience in developing expert and knowledge-based systems for academic and industrial applications in Mexico. The strategy includes KBS (knowledge-based system) technology assimilation, applied research and development with graduate students in computer science and artificial intelligence, an annual program of seminars on expertsystems, an annual international symposium on artificial intelligence applications, research agreements with companies to develop expertsystem prototypes in manufacturing, including the training of knowledge engineers from the sponsoring companies, and the formation of the Center for Artificial Intelligence to do human resources education, applied research, and technological development. Lessons learned in selling expertsystems, problem domain selection, cost/benefit analysis, management of expertsystems projects, and expertsystem tool selection are mentioned.
FAME is a long term project within the IBM Research Division, one of whose objectives is to provide a comprehensive advisory system for financial marketing to IBM marketing representatives. At the same time, FAME has ...
详细信息
ISBN:
(纸本)0818620595
FAME is a long term project within the IBM Research Division, one of whose objectives is to provide a comprehensive advisory system for financial marketing to IBM marketing representatives. At the same time, FAME has served as a test vehicle for a number of ideas relating to knowledge representation, problem solving, and the management of large knowledge based systems. The author describes the FAME expertsystem, discusses its current status, giving particular emphasis to testing and validation procedures that were instituted, and indicates what lessons have been learned. It is concluded that the use of a structured inheritance network, such as K-Rep, as the core knowledge representation service for an expertsystem provides considerable benefits in the construction of an expertsystem. Its enforced semantics encourages a disciplined method of knowledge engineering. This forces system designers to make more careful representation decisions than would otherwise be made in a system without enforced semantics;at the same time, it performs a built-in consistency check as the attributes of objects in K-Rep are assigned values.
The CALTREC project automated the process of creating State of California Travel Expense Claims. This project is a good representative of the types of projects expected to be done by the State of California. The appli...
详细信息
ISBN:
(纸本)0818620595
The CALTREC project automated the process of creating State of California Travel Expense Claims. This project is a good representative of the types of projects expected to be done by the State of California. The application is heavily embedded into traditional data processing systems, which added much additional work outside the scope of what is considered to be knowledge engineering. The author describes the CALTREC application and why it was built, what it took to build the application and why traditional data processing skills were essential to the project's success. It is noted that the present work should be of interest to those who are considering building expertsystems as well as to those who are evaluating alternative software and hardware platforms for implementation and deployment of expertsystems applications.
The authors address the problem faced by expertsystem developers who use microcomputer-based expertsystem shells requiring external interfaces to perform functions, such as communication, mathematical operations and...
详细信息
ISBN:
(纸本)0818620595
The authors address the problem faced by expertsystem developers who use microcomputer-based expertsystem shells requiring external interfaces to perform functions, such as communication, mathematical operations and file handling, that are not provided by the shells. A software development environment, the Data Resource Knowledge Acquisition and Retrieval Tool (DRKART), provides editing support, both for rules comprising an expertsystem's knowledge base and procedural code required for its external interface. It has a built-in communications subsystem and facilities for recording an expert's efforts in accessing and retrieving information from external data sources. The recorded efforts are stored as C functions in DRKART's relational database along with rule constructs required to call them from the knowledge base. These features enable DRKART to provide a simple but novel methodology for constructing microcomputer-based expertsystems.
Since 1982 the US National Association of Corrosion Engineers (NACE) and the US National Institute of Standards and Technology (NIST) have collaborated under the NACE-NIST Corrosion Data Program to develop data resour...
详细信息
ISBN:
(纸本)0818620595
Since 1982 the US National Association of Corrosion Engineers (NACE) and the US National Institute of Standards and Technology (NIST) have collaborated under the NACE-NIST Corrosion Data Program to develop data resources of interest to the materials science and engineering communities. In the last three years these resources have included expertsystems, which have focused primarily on materials selection and usage applications in industry. A brief history of the program is presented, the expertsystems developed therein are presented, and development methodologies are described. Particular emphasis is placed on the participation of the chemical process, oil and gas, and electric power industries in the NACE-NIST Corrosion Data Program.
The Technician's Assister system (TAS) was developed to demonstrate the feasibility for adaptation to the SQS-53B sonar system's diagnostics. The prototype expertsystem incorporates a hierarchical knowledge b...
详细信息
ISBN:
(纸本)0818620595
The Technician's Assister system (TAS) was developed to demonstrate the feasibility for adaptation to the SQS-53B sonar system's diagnostics. The prototype expertsystem incorporates a hierarchical knowledge base design. TAS defines diagnostic testing sequences for system maintenance. Validation of the resulting prototype was performed. TAS provided internal utilities for knowledge base verification. A knowledge-base editor was used to ensure syntactic conformity. The knowledge-base compiler provided diagnostic tests aiding in knowledge accuracy and verified the syntax. A built-in simulator confirmed the data continuity of knowledge base. Data were validated by system-generated rulesets and ambiguity set information. A test-case study of the left correlator database is described. The validation process is divided into three areas: validation of the knowledge base, simulation, and field testing. expert confirmation of the knowledge base and simulation test information was essential for validation. Case simulations were conducted for macromodules and replaceable modules. Consistent results were obtained in the simulated testcase studies.
The management aspects of developing knowledge-based systems software are examined. These management concepts and heuristics have evolved over the past nine years in conjunction with the development of knowledge-based...
详细信息
ISBN:
(纸本)0818620595
The management aspects of developing knowledge-based systems software are examined. These management concepts and heuristics have evolved over the past nine years in conjunction with the development of knowledge-based systems for US Air Force and Army battle managers. However, these development techniques and associated management procedures can also be applied to knowledge-based software systems in general, as well as to software applications which embody operations research and decision analytic technologies. The present discussion of the management methodology uses the example of an ongoing program in artificial intelligence, namely the Heuristic Route Optimization (HERO) program.
The project management, knowledge acquisition, and testing traceability of KBSs (knowledge-based systems) are addressed. Conventional software engineering mechanisms are utilized and woven into a process that encompas...
详细信息
ISBN:
(纸本)0818620595
The project management, knowledge acquisition, and testing traceability of KBSs (knowledge-based systems) are addressed. Conventional software engineering mechanisms are utilized and woven into a process that encompasses the entire development lifecycle. Over the course of several KBS development projects, a process has been created and a suite of tools that assist in its execution has been selected. The projects have ranged from typical interactive-type systems to embedded and flight-worthy code. It was found that knowledge acquisition is key to the KBS lifecycle. A sophisticated tool, gIBIS, was used to combat this problem. This tool provides a visual and expedient methodology for the categorization of data, its representation and relationships, intuitive interface, and high-level prototyping capabilities. Traditional software engineering mechanisms have been coupled to gIBIS to create a flexible environment for the development and management of KBSs.
暂无评论