Mathematical linguistics (Q2426591)

The book deals rather with mathematics inspired by linguistics than with applications of mathematics in linguistics. A large part of the book deals with mathematical models of signal, speech, and handwriting recognition. The book consists of ten chapters. Every chapter contains interesting examples, elegant mathematics and many further reading hints. Chapters 1 and 2 are introductory. They identify prerequisites and present classical topics, like the Chomsky hierarchy. Chapter 3 deals with models of phonology and Chapter 4 discusses morphology, optimality theory and some statistical properties of text bodies, like Zipf's law. The fifth chapter presents several highly abstract syntax models, like categorial grammars, dependency frameworks, weighted finite automata, Markov models, etc. The aspects of syntax parsing are almost completely missing. The semantics models in the sixth chapter are mainly borrowed from mathematical logic. The models include Montague's theory, type theories and grammatical semantics. The models do not include more ``informal'' attitudes like ontologies and other ``real-world'' semantic models. The complexity models in Chapter 7 are based on entropy and Kolmogorov's measure. The last two chapters of the book deal with pattern recognition in signal processing, document classification, speech, and handwriting. The book contains many interesting mathematical models, stimulating examples and elegant proofs, and it is well written. It requires readers having mathematical abilities and a good mathematical background. The book contains almost no topics being important in current computational linguistics, like parsing, maintenance, and analysis of large text bodies (corpuses). It almost does not deal with methods related to dependency grammars that are so important in Slavonic languages and also in English and other non-Slavonic languages.