Partridge - Artificial Intelligence and Software Engineering - Understanding the Promise of the Future.pdf

(2617 KB) Pobierz
ARTIFICIAL INTELLIGENCE
and SOFTWARE ENGINEERING
Understanding the Promise of the Future
Derek Partridge
Glenlake Publishing Company, Ltd.
Chicago • London • New Delhi
AMACOM
American Management Association
New York • Atlanta • Boston • Chicago • Kansas City • San Francisco • Washington, D.C.
Brussels • Mexico City • Tokyo • Toronto
© 1998 Intellect Ltd.
ISBN: 0-8144-0441-3
All rights reserved. No part of this book may be reproduced in any form or by any means, electronic,
mechanical photocopying, recording, or otherwise without the prior written permission of the publisher.
Printed in the United States of America.
AMACOM
American Management Association
1601 Broadway
New York, New York 10019
Visit the American Management Association and AMACOM on-line at
http:\\www.amanet.org
Page iii
Contents
1 Introduction to Computer Software
Computers and software systems
An introduction to software engineering
Bridges and buildings versus software systems
The software crisis
A demand for more software power
Responsiveness to human users
Software systems in new types of domains
Responsiveness to dynamic usage environments
Software systems with self-maintenance capabilities
A need for Al systems
2 AI Problems and Conventional SE Problems
What is an AI problem?
Ill-defined specifications
Correct versus 'good enough' solutions
It's the HOW not the WHAT
1
1
2
4
26
29
29
30
31
32
32
33
33
35
37
38
The problem of dynamics
The quality of modular approximations
Context-free problems
3 Software Engineering Methodology
Specify and verify—the SAV methodology
The myth of complete specification
What is verifiable?
Specify and test—the SAT methodology
40
40
42
45
46
47
54
55
Page iv
Testing for reliability
The strengths
The weaknesses
What are the requirements for testing?
What's in a specification?
Prototyping as a link
4 An Incremental and Exploratory Methodology
Classical methodology and AI problems
56
57
58
59
61
64
71
71
The RUDE cycle
How do we start?
Malleable software
AI muscles on a conventional skeleton
How do we proceed?
How do we finish?
The question of hacking
Conventional paradigms
5 New Paradigms for System Engineering
Automatic programming
Transformational implementation
The "new paradigm" of Balzer, Cheatham and Green
Operational requirements of Kowalski
The POLITE methodology
Towards a Discipline of Exploratory Programming
Reverse engineering
Reusable software
Design knowledge
72
74
75
79
80
85
91
93
101
103
109
113
118
129
137
138
143
153

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin