Akshay Kulkarni, Adarsha Shivananda, Anoosh Kulkarni and V Adithya Krishnan

Applied Recommender Systems with Python

Build Recommender Systems with Deep Learning, NLP and Graph-Based Techniques

Akshay Kulkarni
Bangalore, Karnataka, India
Adarsha Shivananda
Hosanagara tq, Shimoga dt, Karnataka, India
Anoosh Kulkarni
Bangalore, India
V Adithya Krishnan
Navi Mumbai, India
ISBN 978-1-4842-8953-2e-ISBN 978-1-4842-8954-9
https://doi.org/10.1007/978-1-4842-8954-9
© Akshay Kulkarni, Adarsha Shivananda, Anoosh Kulkarni, V Adithya Krishnan 2023
Apress Standard
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Apress imprint is published by the registered company APress Media, LLC, part of Springer Nature.

The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

To our families

Preface

This book is dedicated to data scientists who are starting new recommendation engine projects from scratch but don’t have prior experience in this domain. They can easily learn concepts and gain practical knowledge with this book. Recommendation engines have recently gained a lot of traction and popularity in different domains and have a proven track record for increasing sales and revenue.

This book is divided into eleven chapters. The first section, Chapters 1 and 2, covers basic approaches. The following section, which consists of Chapters 3, 4, 5, and 6, covers popular methods, including collaborative filtering-based, content-based, and hybrid recommendation systems. The next section, Chapters 7 and 8, discusses implementing systems using state-of-the-art machine learning algorithms. Chapters 9, 10, and 11 discuss trending and emerging techniques in recommendation systems.

The code for the implementations in each chapter and the required datasets are available on GitHub at github.com/apress/applied-recommender-systems-python.

To successfully perform all the projects in this book, you need Python 3.x or higher running on any Windows- or Unix-based operating system with a processor of 2.0 GHz or higher and a minimum of 4 GB RAM. You can download Python from Anaconda and leverage a Jupyter notebook for all coding purposes. This book assumes you know Keras basics and how to install machine learning and deep learning basic libraries.

Please upgrade or install the latest versions of all the libraries.

Table of Contents
Chapter 1:​ Introduction to Recommendation Systems1
What Are Recommendation Engines?​2
Recommendation System Types2
Types of Recommendation Engines3
Market Basket Analysis (Association Rule Mining)3
Content-Based Filtering4
Collaborative-Based Filtering7
Hybrid Systems8
ML Clustering9
ML Classification10
Deep Learning11
Rule-Based Recommendation Systems12
Summary19
Chapter 2:​ Market Basket Analysis (Association Rule Mining)21
Implementation22
Data Collection22
Importing the Data as a DataFrame (pandas)22
Cleaning the Data23
Insights from the Dataset24
Free Items and Sales32
Top Ten First Choices42
Frequently Bought Together (MBA)45
Apriori Algorithm Concepts47
Association Rules49
Creating a Function51
Visualization of Association Rules54
Summary62
Chapter 3:​ Content-Based Recommender Systems63
Approach63
Implementation64
Data Collection and Downloading Word Embeddings65
Importing the Data as a DataFrame (pandas)65
Preprocessing the Data66
Similarity Measures71
Build a Model Using CountVectorizer74
Build a Model Using TF-IDF Features76
Build a Model Using Word2vec Features78
Build a Model Using fastText Features80
Build a Model Using GloVe Features82
Build a Model Using a Co-occurrence Matrix85
Summary87
Chapter 4:​ Collaborative Filtering89
Implementation90
Data Collection91
About the Dataset91
Memory-Based Approach93
User-to-User Collaborative Filtering94
KNN-based Approach104
Machine Learning104
Summary110
Chapter 5:​ Collaborative Filtering Using Matrix Factorization, Singular Value Decomposition, and Co-Clustering111
Implementation112
Matrix Factorization, Co-Clustering, and SVD112
Summary128
Chapter 6:​ Hybrid Recommender Systems129
Implementation130
Data Collection130
Data Preparation132
Model Building138
Final Model after Combining the Train and Test Data144
Getting Recommendations145
Summary148
Chapter 7:​ Clustering-Based Recommender Systems149
Approach150
Implementation151
Data Collection and Downloading Required Word Embeddings151
Importing the Data as a DataFrame (pandas)151
Preprocessing the Data153
Exploratory Data Analysis154
Model Building158
Summary174
Chapter 8:​ Classification Algorithm–Based Recommender Systems175
Approach175
Implementation176
Data Collection and Download Word Embeddings177
Importing the Data as a DataFrame (pandas)177
Preprocessing the Data178
Feature Engineering187
Exploratory Data Analysis188
Model Building192
Summary206
Chapter 9:​ Deep Learning–Based Recommender System207
Basics of Deep Learning (ANNs)207
Neural Collaborative Filtering (NCF)208
Implementation210
Data Collection211
Importing the Data as a DataFrame (pandas)211
Data Preprocessing212
Train-Test Split214
Modeling and Recommendations214
Summary220
Chapter 10:​ Graph-Based Recommender Systems221
Implementation223
Summary234
Chapter 11:​ Emerging Areas and Techniques in Recommender Systems235
Real-Time Recommendations235
Conversational Recommendations236
Context-Aware Recommenders237
Multi-task Recommenders238
Joint Representation Learning239
Conclusion240
Index243
About the Authors
Akshay R. Kulkarni

A photograph of Akshay R Kulkarni.

is an artificial intelligence (AI) and machine learning (ML) evangelist and a thought leader. He has consulted several Fortune 500 and global enterprises to drive AI and data science–led strategic transformations. He is a Google developer, an author, and a regular speaker at major AI and data science conferences (including the O’Reilly Strata Data & AI Conference and Great International Developer Summit (GIDS)) . He is a visiting faculty member at some of the top graduate institutes in India. In 2019, he was featured as one of India’s “top 40 under 40” data scientists. In his spare time, Akshay enjoys reading, writing, coding, and helping aspiring data scientists. He lives in Bangalore with his family.
 
Adarsha Shivananda

A photograph of Adarsha Shivananda.

is a data science and MLOps leader. He is working on creating world-class MLOps capabilities to ensure continuous value delivery from AI. He aims to build a pool of exceptional data scientists within and outside organizations to solve problems through training programs. He always wants to stay ahead of the curve. Adarsha has worked extensively in the pharma, healthcare, CPG, retail, and marketing domains. He lives in Bangalore and loves to read and teach data science.
 
Anoosh Kulkarni

A photograph of Anoosh Kulkarni.

is a data scientist and a senior AI consultant. He has worked with global clients across multiple domains to help them solve their business problems using machine learning, natural language processing (NLP), and deep learning. Anoosh is passionate about guiding and mentoring people in their data science journey. He leads data science/machine learning meet-ups and helps aspiring data scientists navigate their careers. He also conducts ML/AI workshops at universities and is actively involved in conducting webinars, talks, and sessions on AI and data science. He lives in Bangalore with his family.
 
V. Adithya Krishnan

A photograph of V Adithya Krishnan.

is a data scientist and MLOps engineer. He has worked with various global clients across multiple domains and helped them to solve their business problems using advanced ML applications. He has experience across multiple fields of AI-ML, including time-series forecasting, deep learning, NLP, ML operations, image processing, and data analytics. Presently, he is working on a state-of-the-art value observability suite for models in production, which includes continuous model and data monitoring along with the business value realized. He presented a paper, “Deep Learning Based Approach for Range Estimation,” written in collaboration with the DRDO, at an IEEE conference. He lives in Chennai with his family.
 
About the Technical Reviewer
Krishnendu Dasgupta

A photograph of Krishnendu Dasgupta.

is co-founder of DOCONVID AI. He is a computer science and engineering graduate with a decade of experience building solutions and platforms on applied machine learning. He has worked with NTT DATA, PwC, and Thoucentric and is now working on applied AI research in medical imaging and decentralized privacy-preserving machine learning in healthcare. Krishnendu is an alumnus of the MIT Entrepreneurship and Innovation Bootcamp and devotes his free time as an applied AI and ML research volunteer for various research NGOs and universities across the world.
 
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.17.150.119