Customer Reviews

Computational Linguistics: An Introduction (Studies in Natural Language Processing)

5 star:		(1)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Grishman offers a readable account of the issues faced by Natural Language Processing research. The discussion is bolstered by examples from some of the more interesting NLP systems. These examples are shallow, permitting a light read and keeping the focus of the book on linguistic issues. The majority of the book covers standard syntactic and semantic parsing techniques, while discourse analysis, anaphora resolution and text generation receive only a quick treatment. Although some attention is paid to alternative approaches, Grisham chooses to phrase most of his discussion in transformational terms.

The most serious weakness of this introduction is its age. Certain assertions, such as the claim that "a finite-state network (regular grammar) is unsuitable" for language processing (p. 18) are seen in hindsight as questionable. Further, the current trend towards statistical analysis is unaddressed.

Warning: some understanding of predicate calculus is necessary to follow the sections on semantic analysis.

The fields of Computer Science and Mathematics allow us to explore the capabilities of the human brain. Part of this exploration includes the understanding of spoken languages processing and the automatic interpretation of these via computational machines. This book gives a solid foundation of the field and allows new researchers to put into perspective all new publications related to compilers design materials. I found a number of confusing concepts in new publications so I needed to buy this book to clear in my mind other author's ideas.

This is a book that I strongly recommend to anyone who is serious about languages, compilers, interpreters, etc.

Stavros Vlachoyannis
CS Graduate Student
Polytechnic University