Analogical Natural Language Processing

The use of examples as the basis for machine translation systems has been a major feature of the last decade's research. Developments in this fertile area are described in "analogical natural language processing" with a thorough discussion of their theoretical and practical implications. The book outlines the fundamental concepts which distinguish example-based (or analogical) processing from the traditional rule-based approach. New concepts such as "cloning" and "recombination" are introduced as processes unique to the new paradigm. Some of the new ideas are demonstrated in a series of practical computational experiments which further elucidate the book's central concepts and framework. This book is intended for postgraduate and research level in machine translation and computational linguistics.