INDIAN INSTITUTE OF MANAGEMENT AMRITSAR

Master of Business Administration

Course Outline

Course Code and Course Title Natural Language Processing

Course type (Regular/Elective) Regular
Pre-requisites (if any) None
Course Units 1
Total no. of sessions 20
Session Duration 75mins
Term IV
Year and Batch 2023, 2022-24 MBA-BA
Sections (if any) 1

Instructor(s) Sunil Reddy Kunduru

Contact Details 9686499792
Office R No 9 New Leased Building
Consultation Hours By appointment

Introduction
Some of the leading firms provide products services to millions of customers every day. The
need for customer engagement is nevertheless very important. Every customer expects the
firms to pay enough attention to their needs and cater to their concerns. As a result of
exponential increase in volume and steady but slow rise in customer expectations, it has
become impossible for firms to engage with customers using human resources alone. There is
a dire need of augmenting human effort in customer engagement with intelligent machine-
based engagement.
Natural Language Processing provides the basic essential tools for machine-based
engagement with increasing volume of customers. It provides tools to automate parsing
through voice and text data to categorize customer messages into service requests,
complaints, queries, etc. It further provides tools to understand, to a limited extent, the
content of the message and to generate replies to the customer messages in such a way that
customer engagement throughout the customer journey is automated.
The need of the hour for businesses across the world is a pool of managers who understand
NLP in some depth so as to lead teams in the direction of automation of customer
engagement. This course is designed to cater to this need of the industry. It provides a
comprehensive understanding of the fundamental of NLP as well as provides ample scope to
practice some of the building blocks of NLP.

Learning Outcomes/Course Objectives

 Understand the building blocks in NLP
 Work with nltk module in Python
 Read and build text corpora
 Extract information from text
 Work with the ambiguities of grammar in English language

Textbooks and Learning Materials

Text Book(s)
Vajjala, S., Majumder, B., Gupta, A., & Surana, H. (2020). Practical natural language
processing: a comprehensive guide to building real-world NLP systems. O'Reilly Media.

Reference Book(s)
Bird, S., Klein, E., and Loper, E. (2009). Natural Language Processing with Python.
O’Reilly.
Christian, B. (2020). The alignment problem: Machine learning and human values. WW
Norton & Company.

Additional Reading(s)
Case 1: Taiger AI: Unbundling the Business Value of NLP By: Seema Chokshi, Lipika
Bhattacharya, Wee-Kiat Lim. Product #: SMU939-PDF-ENG
Case 2: Tamarin App: Natural Language Processing By: Srikant M. Datar, Caitlin N.
Bowler. Product #: 118015-PDF-ENG
Case 3: HubSpot and Motion AI: Chatbot-Enabled CRM By: Jill Avery, Thomas Steenburgh.
Product #: 518067-PDF-ENG
Case 4: Understanding Text Mining and Sentiment Analysis in Hotel Booking By: Rasha
Kashef, Sakariya Ahmed. Product #: W19474-PDF-ENG
Case 5: Blackbox Chatbot: Designing Natural Language Conversations with Data By:
Tamas Makany, Lipika Bhattacharya. Product #: SMU068-PDF-ENG
Case 6: Timnit Gebru: "SILENCED No More" on AI Bias and The Harms of Large
Language Models By: Tsedal Neeley, Stefani Ruper. Product #: 422085-PDF-ENG

Article 1: The Power of Natural Language Processing. By: Ross Gruetzemacher. Product
Number: H06ZS3-PDF-ENG
Article 2: Create Winning Customer Experiences with Generative AI By: Nicolaj Siggelkow;
Christian Terwiesch. Product Number: H07JY6-PDF-ENG
Article 3: Generative AI Won't Revolutionize Search – Yet. By: Ege Gurdeniz; Kartik
Hosanagar. Product Number: H07IBA-PDF-ENG
Article 4: Does Your Company Really Need a Chatbot? By: PV Kannan, Josh Bernoff.
Product #: H04YFZ-PDF-ENG

Technology and Software (if any)

1. Python 3.0
2. Anaconda Navigator
3. Jupyter Notebook

Other Resources (Journals, Internet Websites) (if any)

Pedagogy Used/Learning Process

The course needs three-pronged learning. The students need to understand the nature of text
data, learn to use the tools to analyze the text data and understand the complexities in the
natural language in which the text is produced. Each session will be divided into two parts.
The first part will involve engagement with concepts through discussion of content from
textbook and case discussion as applicable. The second part of each session involves Python
implementation of the concepts discussed in the class. The whole course would be conducted
using modern text editors to facilitate sharing and editing of content between instructor and
students as well as between students.
Evaluation Components/Assessment of Student Learning

Assessment Tool Percentage Description

Quizzes 20 These components are required to
End term Exam 35 check and evaluate the students’
understanding of concepts and
application of appropriate tools and
techniques.
Class Participation (Case 20 Students are expected to work in teams
Analysis & Class Exercises) for case analysis as it would provide
them opportunities to explore diverse
perspectives for a business problem.
Project 25 This facilitates students to demonstrate
the ability to integrate the information
and the skills accumulated in their
course with the applied research in the
respective areas. This also gives them
a chance to show their skills in oral
and writing communication.

Session Plan

Session Module Topic Chapter No. /

Reading
material / Cases
1 Introduction State-of-the-art Article 1
 LLMs Article 2
 Generative AI Article 3
Python Primer Session notes for
 Jupyter notebooks codes
 Functions and datatypes
 List comprehension
 Strings
2 Text mining to NLP TB – Chapter 1
 Text as unstructured data Session notes for
 Information extraction codes
 Text generation and
translation
Regular expressions
 Search
 Findall
 Replace
3 NLP Infrastructure NLP Pipeline TB - Chapter 2
 Pipeline architecture Session notes for
 Components of NLP codes
pipeline – an overview
 Data sources
The nltk module
 Setting up nltk module
 Reading and building
 corpora
 Tokenization of text
4 Working with text data
 Extraction
 Pre-processing
 Feature engineering
Extraction of text data
 Extraction from web
 Extraction from pdf
5 NLP Modelling
 Models
 Evaluation
 Deployment
Pipelines in nltk
 Model definitions
 Pipeline building
6 Text Vectorization TB – Chapter 3
Representations  Bag of words
 TF-IDF
Vectorization in Python
 User-defined functions
 In-built functions
7 Embedding
 Word2Vec and BERT
 GloVe
 GPT
Embedding in Python
 Representation of simple
text
8 Text Classification Classification of Text TB – Chapter 4
 Applications Case 2
 Naïve Bayes for text
classification
Modeling and Evaluation
 Text classification using
Naïve Bayes
9 Classifier Pipeline
 Embedding
 Deep learning
 Interpretation of models
Sentiment analysis in Python
 Cross-evaluation of
classifiers
10 Extracting Information extraction TB – Chapter 5
Information from fundamentals
Text  Structured and unstructured
data - applications
 Chunking
 Exploring text corpora
Chunking in Python
 Developing and evaluating
chunkers
11 IE Pipeline
 General Pipeline
  Key phrase extraction
 Named entity recognition
 Relation extraction
12 Chatbots Chatbots TB – Chapter 6
 Taxonomy Case 3
 Applications Article 4
13 Dialogue systems
 Pipeline
 Components
 LLM chatbots
14 Advanced NLP NLP Topics - I TB – Chapter 7
Topics  Search and information
retrieval
 Topic modeling
 Text summarization
Topic modeling in Python
15 NLP Topics - I
 Recommender system
 Machine translation
 QnA systems
Simple translator using Python
16 Applications Social Media TB - Chapter 8
17 E-commerce TB – Chapter 9
Case 4
18 NLP Process End-to-end view of the NLP Case 5
process TB - Chapter 11
19 NLP and Ethics Ethics in NLP Case 6
 Data bias RB2
 Model bias
20 Project Presentations