Book Image

Natural Language Processing with Python Quick Start Guide

By : Nirant Kasliwal
Book Image

Natural Language Processing with Python Quick Start Guide

By: Nirant Kasliwal

Overview of this book

NLP in Python is among the most sought after skills among data scientists. With code and relevant case studies, this book will show how you can use industry-grade tools to implement NLP programs capable of learning from relevant data. We will explore many modern methods ranging from spaCy to word vectors that have reinvented NLP. The book takes you from the basics of NLP to building text processing applications. We start with an introduction to the basic vocabulary along with a work?ow for building NLP applications. We use industry-grade NLP tools for cleaning and pre-processing text, automatic question and answer generation using linguistics, text embedding, text classifier, and building a chatbot. With each project, you will learn a new concept of NLP. You will learn about entity recognition, part of speech tagging and dependency parsing for Q and A. We use text embedding for both clustering documents and making chatbots, and then build classifiers using scikit-learn. We conclude by deploying these models as REST APIs with Flask. By the end, you will be confident building NLP applications, and know exactly what to look for when approaching new challenges.
Table of Contents (10 chapters)

Putting it all together – the training loop

We now have a shared vocabulary. You have a notional understanding of what terms like layers, model weights, loss function, and optimizer mean. But how do they work together? How do we train them on arbitrary data? We can train them to give us the ability to recognize cat pictures or fraudulent reviews on Amazon.

Here is the rough outline of the steps that occur inside a training loop:

  • Initialize:
    • The network/model weights are assigned random values, usually in the form of (-1, 1) or (0, 1).
    • The model is very far from the target. This is because it is simply executing a series of random transformations.
    • The loss is very high.
  • With every example that the network processes, the following occurs:
    • The weights are adjusted a little in the correct direction
    • The loss score decreases

This is the training loop, which is repeated...