Practical Natural Language Processing: A Comprehensive Guide to Building Real-World NLP Systems

by Sowmya Vajjala, Bodhisattwa Majumder, Anuj Gupta

Many books and courses tackle natural language processing (NLP) problems with toy use cases and well-defined datasets. But if you want to build, iterate, and scale NLP systems in a business setting and tailor them for particular industry verticals, this is your guide. Software engineers and data scientists will learn how to navigate the maze of options available at each step of the journey.

Through the course of the book, authors Sowmya Vajjala, Bodhisattwa Majumder, Anuj Gupta, and Harshit Surana will guide you through the process of building real-world NLP solutions embedded in larger product setups. You'll learn how to adapt your solutions for different industry verticals such as healthcare, social media, and retail.

With this book, you'


Understand the wide spectrum of problem statements, tasks, and solution approaches within NLP
Implement and evaluate different NLP applications using machine learning and deep learning methods
Fine-tune your NLP solution based on your business problem and industry vertical
Evaluate various algorithms and approaches for NLP product tasks, datasets, and stages
Produce software solutions following best practices around release, deployment, and DevOps for NLP systems
Understand best practices, opportunities, and the roadmap for NLP from a business and product leader's perspective

Genres: Computer Science, Artificial Intelligence, Technical, Programming, Nonfiction, Technology, Science

549 pages, Kindle Edition

Published June 17, 2020

Reviews

[Vinayak Hegde · Rating 5 out of 5]

A fairly comprehensive technical book that looks at the practical applications of natural language processing across different Industry verticals. The book is full of practical considerations when building scalable NLP systems that can evolve. It covers not only the different possible libraries for every stage of the NLP processing but also pipelines and tools to bind the different stages together so that they can work in unison. It also covers the rough edges and gotchas where things don't work as well in a fast-evolving field. This advice is valuable as it is often the place where theory meets practice and sometimes does not deliver.

The book takes a deep look at code + data + research + practical application in a holistic way. The abundance of reference links to research papers, video talks, practical implementations, code repositories, and documentation for specific APIs was very welcome. I spent quite a bit of time reading up some of the interesting references and footnotes. The python notebooks are a good addon to the book (Code is at https://github.com/practical-nlp/prac...) so you could read the book and the execute and change and play with the code side-by-side. Overall a great addition to ML/AI/NLP bookshelf. A must-read for any engineer/data scientist working on building NLP systems or wants to know more about the field.

[Vinay Khobragade · Rating 2 out of 5]

This book is not for coders who want to learn how things happen with code.
For me, reading the book was pretty frustrating. I thought that the book will speak more about methods to do tasks in NLP.
However, the book suggests only the following -
1. Building a model from scratch is very time consuming.
2. You need to use heuristics method first to get your job done.
3. You need to do bootstrapping from other APIs.
4. And then it goes on to tell you about different APIs that are provided by Google Cloud, IBM Cloud, AZURE, etc.

I would have loved if the authors would have - "code more and speak less"

Infact the chapters 1,2 and 3 are just fillers. The only chapter where there is some decent amount of code is chapter 4 - Text Classification.

Even in the deployment chapter all the authors care about is theory. And tell you about different libraries and APIs that you can use.

Probably the book is for those who don't know anything about NLP and come from Web development or Management background.

Even for a simple chat bot the authors didn't bother to show how you can build it from scratch but rather promote services such as Dailogflow and RasaNLU.

They have provided a github repository where all of their codes reside. Of course, it doesn't contain what you want to learn. And from any book (tech) you expect that the authors explain the code to you. Otherwise, there are plenty of many more code resources available that will do the task.

[LeoQuiroa · Rating 5 out of 5]

From the level of detail of the problems on the real cases scenarios, you can imagine that the authors have a lot of experience in the field. Besides that, I found very useful the suggestion of libraries and rule of thumb for every step of the ML cycle. Finally, it gives a general idea of where the NLP is moving forward, which is exciting. I found useful chapters (3 ,5, and 11), and not so useful and a little bit repetitive (1,2,6) but overall it is a must-read in the NLP field.

[Kautuk Kundan · Rating 3 out of 5]

This book gives an overview of all the domains which can benefit from NLP however does not have enough hands-on content. The accompanying code repository is good but the code doesn't explain a lot of things in detail.
I wish there was a little more hands-on content about the mentioned practical use cases.

(I read the early release edition)

[Bruce · Rating 4 out of 5]

Not much on Philosophical merits and theories of Natural Language Processing this is an operational reference, and a practical guide for developing in NLP. It was slightly less comprehensive than I was looking for, but useful nonetheless

[Diana · Rating 4 out of 5]

This book present a overview about NLP tasks, give us a start point to init in this field.

[Nigel Thornton · Rating 4 out of 5]

Good overview of the NLP field.

[Ken · Rating 3 out of 5]

Good overview for non-technical managers involved in NLP. Unfortunately, for data scientists involved in creating products, there are more screenshots than code.

[Amirali · Rating 3 out of 5]

Not that much technical. It could have been a a collection of blogposts on medium or towards-data-science.

[Ashley · Rating 3 out of 5]

Okay for a high-level intro to NLP. Could be a third of its length.

[bimri · Rating 3 out of 5]

I found the insights shared in here to be quite applicable to date, although since chatGPT came to the block, most of the core areas NLP cared to solve; have gone the way of dinosaurs 🦕.