Jalaj Thanaki [Thanaki - Python Natural Language Processing

Python

Natural Language Processing

Explore NLP with machine learning and deep learning techniques

Jalaj Thanaki

BIRMINGHAM - MUMBAI

Python Natural Language Processing

Copyright 2017 Packt Publishing

First published: July 2017 Production reference: 1280717

Published by Packt Publishing Ltd.

Livery Place

35 Livery Street

Birmingham B3 2PB, UK.

ISBN 978-1-78712-142-3 www.packtpub.com

Foreword

Data science is rapidly changing the world and the way we do business --be it retail, banking and financial services, publishing, pharmaceutical, manufacturing, and so on. Data of all forms is growing exponentially--quantitative, qualitative, structured, unstructured, speech, video, and so on. It is imperative to make use of this data to leverage all functions-avoid risk and fraud, enhance customer experience, increase revenues, and streamline operations.

Organizations are moving fast to embrace data science and investing a lot into high-end data science teams. Having spent more than 12 years in the BFSI domain, I get overwhelmed with the transition that the BFSI industry has seen in embracing analytics as a business and no longer a support function. This holds especially true for the fin-tech and digital lending world of which Jalaj and myself are a part of.

I have known Jalaj since her college days and am impressed with her exuberance and selfmotivation. Her research skills, perseverance, commitment, discipline, and quickness to grasp even the most difficult concepts have made her achieve success in a short span of 4 years on her corporate journey.

Jalaj is a gifted intellectual with a strong mathematical and statistical understanding and demonstrates a continuous passion for learning the new and complex analytical and statistical techniques that are emerging in the industry. She brings experience to the data science domain and I have seen her deliver impressive projects around NLP, machine learning, basic linguistic analysis, neural networks, and deep learning. The blistering pace of the work schedule that she sets for herself, coupled with the passion she puts into her work, leads to definite and measurable results for her organization.

One of her most special qualities is her readiness to solve the most basic to the most complex problem in the interest of the business. She is an excellent team player and ensures that the organization gains the maximum benefit of her exceptional talent.

In this book, Jalaj takes us on an exciting and insightful journey through the natural language processing domain. She starts with the basic concepts and moves on to the most advanced concepts, such as how machine learning and deep learning are used in NLP.

I wish Jalaj all the best in all her future endeavors.

Sarita Arora

Chief Analytics Officer, SMECorner

Mumbai, India

Preface

The book title, Python Natural Language Processing, gives you a broad idea about the book. As a reader, you will get the chance to learn about all the aspects of natural language processing (NLP) from scratch. In this book, I have specified NLP concepts in a very simple language, and there are some really cool practical examples that enhance your understanding of this domain. By implementing these examples, you can improve your NLP skills. Don't you think that sounds interesting?

Now let me answer some of the most common questions I have received from my friends and colleagues about the NLP domain. These questions really inspired me to write this book. For me, it's really important that all my readers understand why I am writing this book. Let's find out!

Here, I would like answer some of the questions that I feel are critical to my readers. So, I'll begin with some of the questions, followed by the answers. The first question I usually get asked is--what is NLP? The second one is--why is Python mainly used for developing NLP applications? And last but not least, the most critical question is--what are the resources I can use for learning NLP? Now let's look at the answers!

The answer to the first question is that NLP, simply put, is the language you speak, write, read, or understand as a human; natural language is, thus, a medium of communication. Using computer science algorithms, mathematical concepts, and statistical techniques, we try to process the language so machines can also understand language as humans do; this is called NLP.

Now let's answer the second question--why do people mainly use Python to develop NLP applications? So, there are some facts that I want to share with you. The very simple and straightforward thing is that Python has a lot of libraries that make your life easy when you develop NLP applications. The second reason is that if you are coming from a C or C++ coding background, you don't need to worry about memory leakage. The Python interpreter will handle this for you, so you can just focus on the main coding part. Besides, Python is a coder-friendly language. You can do much more by writing just a few lines of codes, compared to other object-oriented languages. So all these facts drive people to use Python for developing NLP and other data science-related applications for rapid prototyping.

The last question is critical to me because I used to explain the previous answers to my friends, but after hearing all these and other fascinating things, they would come to me and say that they want to learn NLP, so what are the resources available? I used to recommend books, blogs, YouTube videos, education platforms such as Udacity and Coursera, and a lot more, but after a few days, they would ask me if there is a single resource in the form of book, blog, or anything that they could use. Unfortunately, for them, my answer was no. At that stage, I really felt that juggling all these resources would always be difficult for them, and that painful realization became my inspiration to write this book.

So in this book, I have tried to cover most of the essential parts of NLP, which will be useful for everyone. The great news is that I have provided practical examples using Python so readers can understand all the concepts theoretically as well as practically. Reading, understanding, and coding are the three main processes that I have followed in this book to make readers lives easier.

What this book covers

Chapter 1 , Introduction , provides an introduction to NLP and the various branches involved in the NLP domain. We will see the various stages of building NLP applications and discuss NLTK installation.

Chapter 2 , Practical Understanding of Corpus and Dataset , shows all the aspects of corpus analysis. We will see the different types of corpus and data attributes present in corpuses.

We will touch upon different corpus formats such as CSV, JSON, XML, LibSVM, and so on. We will see a web scraping example.

Chapter 3 , Understanding Structure of Sentences , helps you understand the most essential aspect of natural language, which is linguistics. We will see the concepts of lexical analysis, syntactic analysis, semantic analysis, handling ambiguities, and so on. We will use NLTK to understand all the concepts practically.

Chapter 4 , Preprocessing , helps you get to know the various types of preprocessing techniques and how you can customize them. We will see the stages of preprocessing such as data preparation, data processing, and data transformation. Apart from this, you will understand the practical aspects of preprocessing.

Chapter 5 , Feature Engineering and NLP Algorithms , is the core part of an NLP application. We will see how different algorithms and tools are used to generate input for machine learning algorithms, which we will be using to develop NLP applications. We will also understand the statistical concepts used in feature engineering, and we will get into the customization of tools and algorithms.