Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Deep learning for NLP and speech recognition
Kamath U., Liu J., Whitaker J., Springer International Publishing, New York, NY, 2019. 621 pp. Type: Book (978-3-030145-95-8)
Date Reviewed: Dec 22 2020

As has happened in other applied domains, deep learning has revolutionized natural language processing (NLP) systems, from speech recognition to neural machine translation. In 2017, Yoav Goldberg provided an easy-to-read overview of the (then) current state of the art [1] based on a survey paper published the previous year in JAIR [2]. Given the fast-moving nature of the field, that nice survey did not include more recent developments such as transformers, which are now covered, albeit cursorily, in the monograph under review.

This Springer monograph, like many books of its type, has three parts: an introduction, a body, and a conclusion. A brief introduction, or exposition, to machine learning and NLP covers the terminology used in both fields. These introductory chapters present the very basics of supervised learning, the most common families of machine learning algorithms, and even some pointers to machine learning theory. Brief histories of deep learning, NLP, and speech recognition highlight some key turning points since the dawn of artificial intelligence (AI). The first 140 pages of the book provide the overall context for the remaining chapters.

The body (another 160 pages) covers the development of deep learning techniques as applied to NLP in general and speech recognition in particular. A chapter introducing the key issues in deep learning algorithms, from regularization to hyperparameter tuning, paves the way for separate chapters where specific architectures are described. Four chapters delve into distributed word representations (that is, word embeddings and so on), convolutional neural networks, recurrent neural networks, and the use of deep neural networks in hidden Markov model (HMM)-based speech recognition systems.

The final chapters address some recent developments and provide a good survey of the current state of the field, from attention mechanisms and memory-augmented networks to end-to-end speech recognition systems, which entirely prescind the HMM present in statistical speech recognition systems. Separate chapters on transfer learning and reinforcement learning complete the overall picture of the ongoing developments in the field.

With respect to other books on the topic, this monograph might not be too friendly for beginners. It is not a traditional textbook in that it does not always explain the reasons behind some decisions, nor provides enough background information to be used without resorting to external sources. On the plus side, it covers many novel research proposals (up to 2018). Moreover, it blends recent developments, available only from research papers, with interesting case studies at the end of every chapter. These case studies, which can be used as class projects, illustrate how ideas can be translated into working code using a handful of open-source software libraries.

In summary, this book might not be helpful as a standalone textbook. Fortunately, nor is it a cookbook with coding recipes. However, it is at times a list of recipes for specific NLP-related problems, without providing further insight. Maybe this is because we cannot explain it, as neural networks are successful in practice yet mainly operate as black boxes.

Graduate students and researchers might find the book useful, to get up to speed on some NLP research topics. More pragmatic engineers might find the case studies entertaining, for example, to play with neural networks on text and speech data (and even try their very own ideas to see how they perform in practice).

More reviews about this item: Amazon

Reviewer:  Fernando Berzal Review #: CR147143 (2105-0105)
1) Goldberg, Y. Neural network methods for natural language processing. Morgan & Claypool, San Rafael, CA, 2017.
2) Goldberg, Y. A primer on neural network models for natural language processing. Journal of Artificial Intelligence Research 57, (2016), 345–420.
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Natural Language Processing (I.2.7 )
 
 
Neural Nets (C.1.3 ... )
 
 
General (I.2.0 )
 
 
Learning (I.2.6 )
 
Would you recommend this review?
yes
no
Other reviews under "Natural Language Processing": Date
Current research in natural language generation
Dale R. (ed), Mellish C. (ed), Zock M., Academic Press Prof., Inc., San Diego, CA, 1990. Type: Book (9780122007354)
Nov 1 1992
Incremental interpretation
Pereira F., Pollack M. Artificial Intelligence 50(1): 37-82, 1991. Type: Article
Aug 1 1992
Natural language and computational linguistics
Beardon C., Lumsden D., Holmes G., Ellis Horwood, Upper Saddle River, NJ, 1991. Type: Book (9780136128137)
Jul 1 1992
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy