- Shopping Bag ( 0 items )
Doody's Review ServiceReviewer: Malgorzata Fort, PhD (University of Pittsburgh)
Description: This book combines works on information retrieval (IR) in biomedicine using natural language processing (NLP) methods for automated knowledge extraction, integration, and application. It addresses the circular process of mining data using NLP tools, reusing it to enrich and correct the existing structures, which in turn are used to further improve information retrieval.
Purpose: Bioinformatics is a hot subject in recent years and the ambitious goal of showing contributions from an NLP perspective as applied to biology and medicine (BioNPL) is quite worthy.
Audience: The book is written for advanced readers. The many formulas and equations may make it too difficult to digest for novices. However, it will appeal to researchers interested in natural language processing, artificial intelligence, and linguistics.
Features: The introductory essay sets the stage by providing information on text mining services for biomedicine. The closing chapter summarizes some of the achievements investigated in other applications described in this book and serves as an opening window on a new arena where NLP services are needed. In between, the chapters are organized in four main sections. Section I is by far the largest since the research on words, concepts and word-to-word relations is well established. Section II concentrates on research that looks at words in context and investigates broader linguistic granularity to overcome ambiguity of terms. Section III offers works on mining techniques which focus more on structures than on words, on the use of neural network architecture to improve retrieval, and the need in clinical domains, taking the reader from the theoretical to the practical applications of mining techniques. Section IV reviews different NLP software developed and used for IR in biomedicine. While editing seems consistent throughout the book, the last chapter and bibliography leave much to desire: language errors, different dates for the same citation in main text and bibliography, multiple font size inconsistencies and duplicate entries for the same citation in the bibliography.
Assessment: Although challenging, this book presents a thoughtful arrangement of works that provide great insight into the field: achievements, current challenges, and the future of dependable automated information retrieval.