ARTIFICIAL INTELLIGENCE and SOFTWARE ENGINEERING(EBOOK)
Derek Partridge
- New Delhi Glenlake Publishing 1998
- xi,274p
1 Introduction to Computer Software 1 Computers and software systems 1 An introduction to software engineering 2 Bridges and buildings versus software systems 4 The software crisis 26 A demand for more software power 29 Responsiveness to human users 29 Software systems in new types of domains 30 Responsiveness to dynamic usage environments 31 Software systems with self-maintenance capabilities 32 A need for Al systems 32 2 AI Problems and Conventional SE Problems 33 What is an AI problem? 33 Ill-defined specifications 35 Correct versus 'good enough' solutions 37 It's the HOW not the WHAT 38 The problem of dynamics 40 The quality of modular approximations 40 Context-free problems 42 3 Software Engineering Methodology 45 Specify and verify—the SAV methodology 46 The myth of complete specification 47 What is verifiable? 54 Specify and test—the SAT methodology 55 Testing for reliability 56 The strengths 57 The weaknesses 58 What are the requirements for testing? 59 What's in a specification? 61 Prototyping as a link 64 4 An Incremental and Exploratory Methodology 71 Classical methodology and AI problems 71 The RUDE cycle 72 How do we start? 74 Malleable software 75 AI muscles on a conventional skeleton 79 How do we proceed? 80 How do we finish? 85 The question of hacking 91 Conventional paradigms 93 5 New Paradigms for System Engineering 101 Automatic programming 103 Transformational implementation 109 The "new paradigm" of Balzer, Cheatham and Green 113 Operational requirements of Kowalski 118 The POLITE methodology 129 Towards a Discipline of Exploratory Programming 137 Reverse engineering 138 Reusable software 143 Design knowledge 153 Stepwise abstraction 156 The problem of decompiling 160 Controlled modification 162 Structured growth 172 Machine Learning: Much Promise, Many Problems 177 Self-adaptive software 177 The promise of increased software power 179 The threat of increased software problems 179 The state of the art in machine learning 181 Practical machine learning examples 194 Multiversion inductive programming 196 8 Expert Systems: The Success Story 201 Expert systems as Al software 201 Engineering expert systems 203 The lessons of expert systems for engineering Al software 208 9 AI into Practical Software 215 Support environments 216 Reduction of effective complexity 218 Moderately stupid assistance 220 An engineering toolbox 229 Self-reflective software 232 Overengineering software 233 10 Summary and What the Future Holds 245 References 253 Index 261