This book takes an empirical approach to language processing, based on applying statistical and other machine-learning algorithms to large corpora.Methodology boxes are included in each chapter. Each chapter is built around one or more worked examples to demonstrate the main idea of the chapter. Covers the fundamental algorithms of various fields, whether originally proposed for spoken or written language to demonstrate how the same algorithm can be used for speech recognition and word-sense disambiguation. Emphasis on web and other practical applications. Emphasis on scientific evaluation. Useful as a reference for professionals in any of the areas of speech and language processing.
Preface
This is an exciting time to be working in speech and language processing. Historically distinct fields (natural language processing, speech recognition, computational linguistics, computational psycholinguistics) have begun to merge. The commercial availability of speech recognition and the need for Web-based language techniques have provided an important impetus for development of real systems. The availability of very large on-line corpora has enabled statistical models of language at every level, from phonetics to discourse. We have tried to draw on this emerging state of the art in the design of this pedagogical and reference work:
Coverage
In attempting to describe a unified vision of speech and language processing, we cover areas that traditionally are taught in different courses in different departments: speech recognition in electrical engineering; parsing, semantic interpretation, and pragmatics in natural language processing courses in computer science departments; and computational morphology and phonology in computational linguistics courses in linguistics departments. The book introduces the fundamental algorithms of each of these fields, whether originally proposed for spoken or written language, whether logical or statistical in origin, and attempts to tie together the descriptions of algorithms from different domains. We have also included coverage of applications like spelling-checking and information retrieval and extraction as well as areas like cognitive modeling. A potential problem with this broad-coverage approach is that it required us to include introductory material for each field; thus linguists may want to skip our description of articulatory phonetics, computer scientists may want to skip such sections as regular expressions, and electrical engineers skip the sections on signal processing. Of course, even in a book this long, we didn't have room for everything. Thus this book should not be considered a substitute for important relevant courses in linguistics, automata and formal language theory, or, especially, statistics and information theory. Emphasis on Practical Applications
It is important to show how language-related algorithms and techniques (from HMMs to unification, from the lambda calculus to transformation-based learning) can be applied to important real-world problems: spelling checking, text document search, speech recognition, Web-page processing, part-of-speech tagging, machine translation, and spoken-language dialogue agents. We have attempted to do this by integrating the description of language processing applications into each chapter. The advantage of this approach is that as the relevant linguistic knowledge is introduced, the student has the background to understand and model a particular domain. Emphasis on Scientific Evaluation
The recent prevalence of statistical algorithms in language processing and the growth of organized evaluations of speech and language processing systems has led to a new emphasis on evaluation. We have, therefore, tried to accompany most of our problem domains with a Methodology Box describing how systems are evaluated (e.g., including such concepts as training and test sets, cross-validation, and information-theoretic evaluation metrics like perplexity). Description of widely available language processing resources
Modern speech and language processing is heavily based on common resources: raw speech and text corpora, annotated corpora and treebanks, standard tagsets for labeling pronunciation, part-of-speech, parses, word-sense, and dialogue-level phenomena. We have tried to introduce many of these important resources throughout the book (e.g., the Brown, Switchboard, callhome, ATIS, TREC, MUC, and BNC corpora) and provide complete listings of many useful tagsets and coding schemes (such as the Penn Treebank, CLAWS C5 and C7, and the ARPAbet) but some inevitably got left out. Furthermore, rather than include references to URLs for many resources directly in the textbook, we have placed them on the book's Web site, where they can more readily updated.
The book is primarily intended for use in a graduate or advanced undergraduate course or sequence. Because of its comprehensive coverage and the large number of algorithms, the book is also useful as a reference for students and professionals in any of the areas of speech and language processing. Overview of the Book
The book is divided into four parts in addition to an introduction and end matter. Part I, "Words", introduces concepts related to the processing of words: phonetics, phonology, morphology, and algorithms used to process them: finite automata, finite transducers, weighted transducers, N-grams, and Hidden Markov Models. Part II, "Syntax", introduces parts-of-speech and phrase structure grammars for English and gives essential algorithms for processing word classes and structured relationships among words: part-of-speech taggers based on HMMs and transformation-based learning, the CYK and Earley algorithms for parsing, unification and typed feature structures, lexicalized and probabilistic parsing, and analytical tools like the Chomsky hierarchy and the pumping lemma. Part III, "Semantics", introduces first order predicate calculus and other ways of representing meaning, several approaches to compositional semantic analysis, along with applications to information retrieval, information extraction, speech understanding, and machine translation. Part IV, "Pragmatics", covers reference resolution and discourse structure and coherence, spoken dialogue phenomena like dialogue and speech act modeling, dialogue structure and coherence, and dialogue managers, as well as a comprehensive treatment of natural language generation and of machine translation. Using this Book
The book provides enough material to be used for a full-year sequence in speech and language processing. It is also designed so that it can be used for a number of different useful one-term courses:
NLP
1 quarter NLP
1 semester Speech + NLP
1 semester Comp. Linguistics
1 quarter
1. Intro 1. Intro 1. Intro1. Intro
2. Regex, FSA 2. Regex, FSA 2. Regex, FSA2. Regex, FSA
8. POS tagging 3. Morph., FST 3. Morph., FST3. Morph., FST
9. CFGs 6. N-grams 4. Comp. Phonol.4. Comp. Phonol.
10. Parsing 8. POS tagging 5. Prob. Pronun.10. Parsing
11. Unification 9. CFGs 6. N-grams11. Unification
14. Semantics 10. Parsing 7. HMMs & ASR13. Complexity
15. Sem. Analysis 11. Unification 8. POS tagging16. Lex. Semantics
18. Discourse 12. Prob. Parsing 9. CFGs18. Discourse
20. Generation 14. Semantics 10. Parsing19. Dialogue
15. Sem. Analysis 12. Prob. Parsing
16. Lex. Semantics 14. Semantics
17. WSD and IR 15. Sem. Analysis
18. Discourse 19. Dialogue
20. Generation 21. Mach. Transl.
21. Mach. Transl.
Selected chapters from the book could also be used to augment courses in Artificial Intelligence, Cognitive Science, or Information Retrieval.
