This book is about knowledge representation. Knowledge representation has emerged as one of the fundamental topics in the area of computer science popularly known as artificial intelligence (AI). The reason for this is simple: the basic idea is that intelligence, whatever else it involves, does at least involve knowing things, and exploiting them so as to respond appropriately to a given situation. Therefore it seems reasonable to suppose that if we wish to make an intelligent computer system we must have a way of getting it to know things, and that involves finding a way of representing the things we wish it to know so that they can be encoded within the computer system. Thus the intention of the book is to provide an introduction to the field of artificial intelligence, via an examination of issues concerning knowledge representation. In the first chapter, we consider what is meant by "artificial intelligence", and the related terms like knowledge-based systems and expert systems, so that we can be clear about our motivation, and about exactly what we are trying to achieve with such programs. The second chapter gives a general consideration of what knowledge representation is, and what criteria we can use to access a particular form of knowledge representation. The third chapter outlines some of the basic concepts and terminology of formal logic, which is essential as a means of describing and distinguishing between the various computer-oriented paradigms discussed later. The fourth chapter gives a description of the key topics relating to search, which is historically important in AI, and which is also crucial to the understanding of a range of knowledge representation issues. Chapters 5, 6 and 7 consider the three leading representation paradigms - - production rules, structured object representations and predicate calculus, - so that we can see the style of systems they produce, and consider their various strengths and weaknesses. To illustrate some of the use that has been made of these paradigms the eighth chapter examines PROLOG, which can be seen as a practical instantiation of many of the ideas deriving from the predicate logic paradigm, and the ninth chapter discusses expert systems, which represent an area of application which has made extensive use both of production rules and structured object representations. In the last chapter a number of issues relating to the use and limitations of the paradigms are discussed, so as to indicate some of the unsolved problems of knowledge representation
Description:
This book is about knowledge representation. Knowledge representation has emerged as one of the fundamental topics in the area of computer science popularly known as artificial intelligence (AI). The reason for this is simple: the basic idea is that intelligence, whatever else it involves, does at least involve knowing things, and exploiting them so as to respond appropriately to a given situation. Therefore it seems reasonable to suppose that if we wish to make an intelligent computer system we must have a way of getting it to know things, and that involves finding a way of representing the things we wish it to know so that they can be encoded within the computer system. Thus the intention of the book is to provide an introduction to the field of artificial intelligence, via an examination of issues concerning knowledge representation. In the first chapter, we consider what is meant by "artificial intelligence", and the related terms like knowledge-based systems and expert systems, so that we can be clear about our motivation, and about exactly what we are trying to achieve with such programs. The second chapter gives a general consideration of what knowledge representation is, and what criteria we can use to access a particular form of knowledge representation. The third chapter outlines some of the basic concepts and terminology of formal logic, which is essential as a means of describing and distinguishing between the various computer-oriented paradigms discussed later. The fourth chapter gives a description of the key topics relating to search, which is historically important in AI, and which is also crucial to the understanding of a range of knowledge representation issues. Chapters 5, 6 and 7 consider the three leading representation paradigms - - production rules, structured object representations and predicate calculus, - so that we can see the style of systems they produce, and consider their various strengths and weaknesses. To illustrate some of the use that has been made of these paradigms the eighth chapter examines PROLOG, which can be seen as a practical instantiation of many of the ideas deriving from the predicate logic paradigm, and the ninth chapter discusses expert systems, which represent an area of application which has made extensive use both of production rules and structured object representations. In the last chapter a number of issues relating to the use and limitations of the paradigms are discussed, so as to indicate some of the unsolved problems of knowledge representation