Publisher: Cambridge University Press
Pages: 405, Date: 2010-10-31
ISBN-10 : 0521757606, rared HQ PDF, 1.51 MB
Computational semantics is the art and science of computing meaning in natural language. The meaning of a sentence is derived from the meanings of the individual words in it, and this process can be made so precise that it can be implemented on a computer.
Designed for students of linguistics, computer science, logic and philosophy, this comprehensive text shows how to compute meaning using the functional programming language Haskell.
It deals with both denotational meaning (where meaning comes from knowing the conditions of truth in situations), and operational meaning (where meaning is an instruction for performing cognitive action).
Including a discussion of recent developments in logic, it will be invaluable to linguistics students wanting to apply logic to their studies, logic students wishing to learn how their subject can be applied to linguistics, and functional programmers interested in natural language processing as a new application area.
The book starts without any presuppositions. Nothing is assumed beyond common sense: no programming skills, no knowledge of syntax,no knowledge of semantics, no knowledge of natural language processing of any sort. The book begins by introducing the basics of Haskell, the programming language that is used throughout.
Haskell is a member of the family of functional programming languages and suitable for the theoretical purposes of this book because of its exceptional transparency. The use of Haskell in an introduction to computational semantics is a departure from the widespread use of Prolog in introductions to symbolic natural language processing.
The authors motivate this choice by pointing out that much of symbolic computational linguistics consists, like so many other types of computational problems, in defining the right data structures and the right functions that map one data structure (the one that presents the problem) into another (the one that presents its solution). Indeed, the first applications that are shown in the book are non-linguistic, and chosen solely for the purpose of giving the reader a good grasp of how Haskell works and what can be done with it. From these first applications there is a gradual progress, via applications that involve elementary forms of language processing (such as word counts in texts), to applications that belong to computational semantics proper. One important benefit of this way of easing into the subject is that it makes the reader aware of how much language processing has in common with problem solving and data processing tasks that do not involve language...
Although the book starts from scratch, and presupposes nothing beyond common sense, it nevertheless manages to penetrate into some of the more advanced areas of computational semantics: the final chapters present continuation semantics, discourse representation theory, and a system of dynamic epistemic logic that serves as a first step in the direction of a full-fledged computational theory of verbal communication.
And yet, this wide span notwithstanding, the book complies with current demands on texts in computational linguistics in that it presents for every bit of theory that it introduces a Haskell implementation, which the reader can run herself and play with, if she downloads the relevant, freely available supporting software. The same goes for the numerous implementation-related exercises.
This means that the diligent student of this book will, by the time she gets to the end of it, have not only learned a good deal about syntax and semantics, but have also acquired much of that which distinguishes the computational from the theoretical linguist.
Download Link
.
0 comments:
Post a Comment