Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!

Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


3.6. CASED-BASED REASONING TOOLS

Cased-based reasoning (CBR) tools are somewhat recent additions to the toolbox of the knowledge engineer. In practice, they serve a role similar to induction tools in that they use past experiences (cases) to solve current problems. Given an input specification of a problem, the system will search its case memory for an existing case that matches the input specification. It may find an exact match and immediately go to a solution. Even if an exact match can't be found, the system applies a matching algorithm in order to find a case that is most similar to the input specification.

A CBR tool shares many of the same benefits found in an induction tool, such as being a good choice if a large set of examples exist, and being applicable for a problem where no real expert exists. A case-based expert system can also be built quickly and easily maintained by adding new cases.

3.7. DOMAIN-SPECIFIC TOOLS

The early shells were general-purpose tools and were offered to address a broad range of problem-solving activities. In many instances, however, it was found that though the tool was fine for one activity (e.g., diagnosis), it was inadequate for another one (e.g., design). To address this situation, the AI community, including both software vendors and researchers, turned their attention to developing domain-specific tools. A domain-specific tool is designed to be used to develop an expert system for a particular problem-solving activity. It provides special features to the developer that are tailored for producing an expert system for the activity. Table 1 shows the typical types of problems addressed by expert system developers (adapted from Hayes-Roth et al., 1983). The following sections describe some of the domain-specific tools developed for most of these activities. Some are commercially available while others were developed by organizations to satisfy their particular needs.

Control. ASIA assists with building real-time control expert systems (Devedzic and Velasevic, 1992). The system is capable of reasoning with external real-time data, using symbolic processing. It was developed at Mihailo Pupin Institute, Belgrade, Yugoslavia.

FAIN, Fast AI shell of Nippon Steel, was developed by Nippon Steel for the design of an expert system for a control application (Wakisaka et al., 1993). The expert system is developed by adding or revising the design specifications in a spiral-up manner. The design document specification suited for the prototyping method is specified and the program is automatically generated from the specification.

G2 is a graphical object-oriented environment for building intelligent process management solutions. Its natural language editor allows users to enter rules, models, and procedures that describe real-time operations. Also available are a number of add-on packages that run on top of G2, including scheduling systems, fuzzy logic, diagnostic packages, genetic algorithms, and neural networks. Typical applications built using G2 include process optimization, real-time quality management, supervisory control, and advanced control using fuzzy logic and neural network techniques. G2 is commercially available from Gensym, Cambridge, MA.

An intelligent control shell for CAD tools was developed that can automatically create a command sequence to control CAD systems using symbolic knowledge of general command flows and nonsymbolic knowledge of the past execution data (Fujita et al., 1994). Users define a model of possible control flows, which are transformed into a state transition graph from which executable command sequences are inferred. The control system statistically analyzes nondeterministic branches, where a final result is predicted from a current state of a design object, a command history, and the succeeding commands. Then, the most promising command to optimize the design objects is selected and executed. It was developed at NEC Corp., Kanagawa, Japan.


TABLE 1
Types of Problems Solved by Expert Systems
 
Problem-solving activity Description
Control Governing system behavior to meet specifications
Design Configuring objects under constraint
Diagnosis Inferring system malfunctions from observables
Instruction Diagnosing, debugging, and repairing student behavior
Interpretation Inferring situation description from data
Monitoring Comparing observations to expectations
Planning Selecting and sequencing activities according to a set of constraints to achieve a goal
Prediction Inferring likely consequences of given situations
Prescription Recommending solution to system malfunction
Scheduling Assigning resources and times to the set of activities in a plan
Selection Identifying best choice from a list of possibilities
Simulation Modeling the interaction between system components

A shell was developed for robotics and flexible manufacturing systems (FMS) applications (Devedzic and Krtolica, 1989). The shell is intended for building and testing knowledge-based systems in various robotics and FMS domains, like design, planning, control, scheduling, routing, diagnostics, etc. The shell can be used for building and testing knowledge-based systems for real-time operation, which is typical for robotics and FMS control tasks. The shell was developed at Mihailo Pupin Inst., Belgrade, Yugoslavia.


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.