Principles of Artificial Intelligence

Front Cover
Springer Science & Business Media, May 1, 1982 - Computers - 476 pages
Previous treatments of Artificial Intelligence (AI) divide the subject into its major areas of application, namely, natural language processing, automatic programming, robotics, machine vision, automatic theorem proving, intelligent data retrieval systems, etc. The major difficulty with this approach is that these application areas are now so extensive, that each could, at best, be only superficially treated in a book of this length. Instead, I have attempted here to describe fundamental AI ideas that underlie many of these applications. My organization of these ideas is not, then, based on the subject matter of their application, but is, instead, based on general computational concepts involving the kinds of data structures used, the types of operations performed on these data struc tures, and the properties of con'trol strategies used by AI systems. I stress, in particular, the important roles played in AI by generalized production systems and the predicate calculus. The notes on which the book is based evolved in courses and seminars at Stanford University and at the University of Massachusetts at Amherst. Although certain topics treated in my previous book, Problem solving Methods in Artificial Intelligence, are covered here as well, this book contains many additional topics such as rule-based systems, robot problem-solving systems, and structured-object representations.

From inside the book

Contents

PRODUCTION SYSTEMS AND AI
17
12 SPECIALIZED PRODUCTION SYSTEMS
35
13 COMMENTS ON THE DIFFERENT TYPES OF PRODUCTION SYSTEMS
47
14 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
48
EXERCISES
50
SEARCH STRATEGIES FOR AI PRODUCTION SYSTEMS
53
21 BACKTRACKING STRATEGIES
55
22 GRAPHSEARCH STRATEGIES
61
66 CONTROL KNOWLEDGE FOR RULEBASED DEDUCTION SYSTEMS
257
67 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
267
EXERCISES
270
BASIC PLANGENERATING SYSTEMS
275
72 A FORWARD PRODUCTION SYSTEM
281
73 A REPRESENTATION FOR PLANS
282
74 A BACKWARD PRODUCTION SYSTEM
287
75 STRIPS
298

23 UNINFORMED GRAPHSEARCH PROCEDURES
68
24 HEURISTIC GRAPHSEARCH PROCEDURES
72
25 RELATED ALGORITHMS
88
26 MEASURES OF PERFORMANCE
91
EXERCISES
96
SEARCH STRATEGIES FOR DECOMPOSABLE PRODUCTION SYSTEMS
99
A HEURISTIC SEARCH PROCEDURE FOR ANDOR GRAPHS
103
DECOMPOSABLE AND COMMUTATIVE SYSTEMS
109
34 SEARCHING GAME TREES
112
35 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
127
THE PREDICATE CALCULUS IN AI
131
42 RESOLUTION
145
43 THE USE OF THE PREDICATE CALCULUS IN AI
152
44 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
156
RESOLUTION REFUTATION SYSTEMS
161
51 PRODUCTION SYSTEMS FOR RESOLUTION REFUTATIONS
163
52 CONTROL STRATEGIES FOR RESOLUTION METHODS
164
53 SIMPLIFICATION STRATEGIES
172
54 EXTRACTING ANSWERS FROM RESOLUTION REFUTATIONS
175
55 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
189
RULEBASED DEDUCTION SYSTEMS
193
61 A FORWARD DEDUCTION SYSTEM
196
62 A BACKWARD DEDUCTION SYSTEM
212
63 RESOLVING WITHIN ANDOR GRAPHS
234
64 COMPUTATION DEDUCTIONS AND PROGRAM SYNTHESIS
241
65 A COMBINATION FORWARD AND BACKWARD SYSTEM
253
76 USING DEDUCTION SYSTEMS TO GENERATE ROBOT PLANS
307
77 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
315
EXERCISES
317
ADVANCED PLANGENERATING SYSTEMS
321
82 DCOMP
333
83 AMENDING PLANS
342
84 HIERARCHICAL PLANNING
349
85 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
357
EXERCISES
358
STRUCTURED OBJECT REPRESENTATIONS
361
91 FROM PREDICATE CALCULUS TO UNITS
362
SEMANTIC NETWORKS
370
93 MATCHING
378
94 DEDUCTIVE OPERATIONS ON STRUCTURED OBJECTS
387
95 DEFAULTS AND CONTRADICTORY INFORMATION
408
96 BIBLIOGRAPHICAL AND HISTORICAL REMARKS
412
EXERCISES
414
PROSPECTUS
417
101 AI SYSTEM ARCHITECTURES
418
102 KNOWLEDGE ACQUISITION
419
103 REPRESENTATIONAL FORMALISMS
422
BIBLIOGRAPHY
429
AUTHOR INDEX
467
SUBJECT INDEX
471
Copyright

Other editions - View all

Common terms and phrases

Popular passages

Page 6 - One of the important contributions of research in automatic programming has been the notion of debugging as a problem-solving strategy. It has been found that it is often much more efficient to produce an...
Page 7 - These efforts were directed at making the time-versus-problem-size curve grow as slowly as possible, even when it must grow exponentially. Several methods have been developed for delaying and moderating the inevitable combinatorial explosion. Again, knowledge about the problem domain is the key to more efficient solution methods. Many of the methods developed to deal with combinatorial problems are also useful on other, less combinatorially severe problems.
Page 2 - ... It has been very difficult to develop computer systems capable of generating and "understanding" even fragments of a natural language, such as English. One source of the difficulty is that language has evolved as a communication medium between intelligent beings. Its primary use is for transmitting a bit of "mental structure...
Page 5 - It has led to several techniques for modeling states of the world and for describing the process of change from one world state to another. It has led to a better understanding of how to generate plans for action sequences and how to monitor the execution of these plans. Complex robot control problems have forced...
Page 5 - Robotics The problem of controlling the physical actions of a mobile robot might not seem to require much intelligence. Even small children are able to navigate successfully through their environment and to manipulate items, such as light switches, toy blocks, eating utensils, etc. However these same tasks, performed almost unconsciously by humans, when performed by a machine require many of the same abilities used in solving more intellectually demanding problems. Research on robots or robotics...
Page 4 - Many expert consulting systems employ the AI technique of rule-based deduction. In such systems, expert knowledge is represented as a large set of simple rules, and these rules are used to guide the dialogue between the system and the user and to deduce conclusions.

About the author (1982)

Nils John Nilsson was born in Saginaw, Michigan on February 6, 1933. He received a bachelor's degree and a Ph.D. in electrical engineering from Stanford University. He joined the Air Force and served three years at the Rome Air Development Center before joining the Stanford Research Institute in 1961. He helped develop the first general-purpose robot and was a co-inventor of algorithms that made it possible for the machine to move about efficiently and perform simple tasks. He later worked on the "Computer-based Consultant," which focused on natural language understanding. He wrote several books including Learning Machines: Foundations of Trainable Pattern-Classifying Systems and The Quest for Artificial Intelligence: A History of Ideas and Achievements. He died on April 21, 2019 at the age of 86.

Bibliographic information