Blurb: Natural language understanding is the most comprehensive, in-depth book to date covering all major concepts of natural language processing. Written by James Allen, a leading authority in the field of artificial intelligence, this definitive work addresses modern techniques of syntactic processing and the first detailed study of semantic interpretation. It is a comprehensive reference for professionals in the fields of artificial intelligence, expert systems, and linguistics, and is also appropriate for students of computer science, linguistics, and engineering. Features: - State-of-the-art artificial intelligence issues including knowledge representation and world knowledge. - A comprehensive overview of the field that includes both syntax and semantics. - The first book to thoroughly cover semantic issues. - A synthesis of all important techniques, research, and concepts in the field. About the author: James Allen is editor-in-chief of the journal Computational Linguistics, the leading international publication for work in computational linguistics. He is involved in some of the most important research on natural language processing and has published numerous papers in the area of contextual interpretation, including speech act analysis and discourse structure. In addition, he has written papers addressing issues in logical form, specifically those relating to action verbs and tenses. An instructor for ten years, Allen is currently Associate Professor in the Computer Science Department at the University of Rochester. He has taught courses in natural language processing, artificial intelligence programming techniques, and has held advanced seminars in natural language understanding and knowledge representation. In 1984, James Allen was one of four artificial intelligence researchers nationwide to receive the prestigious Presidential Young Investigator Award (flaptekst).

I purchased this more at less at random when I was first investigating the field of natural language processing and haven't looked at it much since. The focus is on deterministic methods, e.g. parsing, unification, and semantic data structures. There is nothing about probabilistic methods. May be a decent reference for the non-stochastic side of things; I really haven't read enough to know.

Nice book, and the best in its day. Heavy on syntax for my taste (about half the book if I recall) but I'm more interested in semantics and knowledge representation anyway. And this is a textbook, so it ought to cover everything evenly: and it does. ( )