Practical Natural Language Processing with Python: With Case Studies from Industries Using Text Data at Scale

Table of Contents
About the Author
About the Technical Reviewer
Acknowledgments
Introduction
Chapter 1: Types of Data
	Search
	Reviews
	Social Media Posts/Blogs
	Chat Data
		Personal Chats
		Business Chats and Voice Call Data
	SMS Data
	Content Data
	IVR Utterance Data
	Useful Information from Data
Chapter 2: NLP in Customer Service
	Voice Calls
	Chats
	Tickets Data
	Email data
	Voice of Customer
		Intent Mining
		Top Words to Understand Intents
		Word Cloud
		Rules to Classify Topics
	Supervised Learning Using Machine Learning
		Getting Manually Labelled Data
		Word Tokenization
		Term-Document Matrix
		Data Normalization
	Replacing Certain Patterns
	Identifying Issue Lines
	Top Customer Queries
	Top CSAT Drivers
	Top NPS Drivers
	Insights into Sales Chats
		Top Products for Sales Chats
		Reasons for Non-Purchase
		Survey Comments Analysis
		Mining Voice Transcripts
		Acoustic Model
		Language Model
Chapter 3: NLP in Online Reviews
	Sentiment Analysis
	Emotion Mining
	Approach 1: Lexicon-Based Approach
	Approach 2: Rules-Based Approach
		Observation 1
		Observation 2
		Observation 3
		Observation 4
		Overall Score
		Implementing the Observations
			Preprocessing
			Booster and Negation Words (Observation 2 and Observation 3)
			Exclamation Marks (Observation 4)
			Evaluation of full_txt and Summary (Observation 1)
			Optimizing the Code
		Sentiment Analysis Libraries
	Approach 3: Machine-Learning Based Approach (Neural Network)
		Corpus Features
			Building the Neural Network
		Things to Improve
	Attribute Extraction
		Step 1: Using Regex to Normalize
		Step 2: Extracting Noun Forms
		Step 3: Creating a Mapping File
		Step 4: Mapping Each Review to an Attribute
		Step 5: Analyzing Brands
Chapter 4: NLP in Banking, Financial Services, and Insurance (BFSI)
	NLP in Fraud
		Method 1: Using Existing Libraries
		Method 2: Extracting Noun Phrases
		Method 3: Training Your Own Model
			Word Embeddings
			Word2Vec
			CBOW
			Other word2vec Libraries
			Applying Word Embeddings to Supervised Learning
				Method 3 – Approach 1
				Method 3 – Approach 2
	Applying the Model
	Other Use Cases of NLP in Banking, Financial Services, and Insurance
		SMS Data
		Natural Language Generation in Banks
Chapter 5: NLP in Virtual Assistants
	Types of Bots
	The Classic Approach
		Quick Overview of LSTM
			Forget Gate
			Input Gate
			Output Gate
		Applying LSTM
		Time-Distributed Layer
	Approach 2 - The Generating Responses Approach
		Encoder-Decoder Architecture
		Dataset
		Implementing the Architecture
		Encoder-Decoder Training
		Encoder Output
		Decoder Input
		Preprocessing
		Bidirectional LSTM
			Encoder
			Decoder
	BERT
		Language Models and Fine-Tuning
		Overview of BERT
		Fine-Tuning BERT for a Classifier
	Further Nuances in Building Conversational Bots
		Single-Turn vs. Multi-Turn Conversations
		Multi-Lingual Bots
Index