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

ISBN-13: 9781484289532 / Angielski / Miękka / 2022 / 248 str.

Akshay Kulkarni; Adarsha Shivananda; Anoosh Kulkarni

Opis

This book will teach you how to build recommender systems with machine learning algorithms using Python. Recommender systems have become an essential part of every internet-based business today.You'll start by learning basic concepts of recommender systems, with an overview of different types of recommender engines and how they function. Next, you will see how to build recommender systems with traditional algorithms such as market basket analysis and content- and knowledge-based recommender systems with NLP. The authors then demonstrate techniques such as collaborative filtering using matrix factorization and hybrid recommender systems that incorporate both content-based and collaborative filtering techniques. This is followed by a tutorial on building machine learning-based recommender systems using clustering and classification algorithms like K-means and random forest. The last chapters cover NLP, deep learning, and graph-based techniques to build a recommender engine. Each chapter includes data preparation, multiple ways to evaluate and optimize the recommender systems, supporting examples, and illustrations.By the end of this book, you will understand and be able to build recommender systems with various tools and techniques with machine learning, deep learning, and graph-based algorithms.

What You Will Learn

• Understand and implement different recommender systems techniques with Python
• Employ popular methods like content- and knowledge-based, collaborative filtering, market basket analysis, and matrix factorization 
• Build hybrid recommender systems that incorporate both content-based and collaborative filtering
• Leverage machine learning, NLP, and deep learning for building recommender systems

Who This Book Is For

Data scientists, machine learning engineers, and Python programmers interested in building and implementing recommender systems to solve problems.

Spis treści

Chapter 1: Introduction to Recommender Systems

Chapter Goal: Introduction of recommender systems, along with a high-level overview of how recommender systems work, what are the different existing types, and how to leverage basic and advanced machine learning techniques to build these systems.

No of pages: 25

Sub - Topics:  

1. Intro to recommender system 

2. How it works

3. Types and how they work

a. Association rule mining

b. Content based

c. Collaborative filtering 

d. Hybrid systems

e. ML Clustering based

f. ML Classification based

g. Deep learning and NLP based

h. Graph based

Chapter 2: Association Rule Mining

Chapter Goal: Building one of the simplest recommender systems from scratch, using association rule mining; also called market basket analysis.

No of pages: 20

Sub - Topics

1 APRIORI

2 FP GROWTH

3 Advantages and Disadvantages

Chapter 3: Content and Knowledge-Based Recommender System

Chapter Goal: Building the content and knowledge-based recommender system from scratch using both product content and demographics

No of pages: 25

Sub - Topics

1 TF-IDF

2 BOW

3 Transformer based

4 Advantages and disadvantages

Chapter 4: Collaborative Filtering using KNN

Chapter Goal: Building the collaborative filtering using KNN from scratch, both item-item and user-user based

No of pages: 25

Sub - Topics: 

1 KNN – item based

2 KNN – user based

3 Advantages and disadvantages

Chapter 5: Collaborative Filtering Using Matrix Factorization, SVD and ALS.

Chapter Goal: Building the collaborative filtering using SVM from scratch, both item-item and user-user based

No of pages: 25

Sub - Topics: 

1 Latent factors

2 SVD

3 ALS

4 Advantages and disadvantages

Chapter 6: Hybrid Recommender System

Chapter Goal: Building the hybrid recommender system (Using both content and collaborative methods) which is widely used in the industry

No of pages: 25

Sub - Topics: 

1 Weighted: a different weight given to the recommenders of each technique used to favor some of them.

2 Mixed: a single set of recommenders, without favorites.

3 Augmented: suggestions from one system are used as input for the next, and so on until the last one.

4 Switching: Choosing a random method

5 Advantages and disadvantages

Chapter 7: Clustering Algorithm-Based Recommender System

Chapter Goal: Building the clustering model for recommender systems.

No of pages: 25

Sub - Topics: 

1 K means clustering

2 Hierarchal clustering 

3 Advantages and disadvantages

Chapter 8: Classification Algorithm-Based Recommender System

Chapter Goal: Building the classification model for recommender systems.

No of pages: 25

Sub - Topics: 

1 Buying propensity model

2 Logistic regression

3 Random forest

4 SVM

5 Advantages and disadvantages

Chapter 9: Deep Learning and NLP Based Recommender System

Chapter Goal: Building state of art recommender system using advanced topics like Deep learning along with NLP (Natural Language processing).

No of pages: 25

Sub - Topics: 

1 Word embedding’s

2 Deep neural networks

3 Advantages and disadvantages

Chapter 10: Graph-Based Recommender System

Chapter Goal: Implementing graph-based recommender system using Python for computation performance

No of pages: 25

Sub - Topics: 

1 Generating nodes and edges

2 Building algorithm

3 Advantages and disadvantages

Chapter 11: Emerging Areas and Techniques in Recommender System 

Chapter Goal: To get an overview of the new and emerging techniques and the areas of research in Recommender systems

No of pages: 15

Sub - Topics: 

1 Personalized recommendation engine

2 Context-based search engine

3 Multi-objective recommendations

4 Summary

O autorze

Akshay R Kulkarni is an AI and machine learning 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, Author, and a regular speaker at major AI and data science conferences including Strata, O’Reilly AI Conf, and GIDS. He is a visiting faculty member for some of the top graduate institutes in India. In 2019, he has been also featured as one of the top 40 under 40 Data Scientists in India. In his spare time, he enjoys reading, writing, coding, and helping aspiring data scientists. He lives in Bangalore with his family.

Adarsha Shivananda is 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 of the organization to solve problems through training programs, and always wants to stay ahead of the curve. He 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 is a data scientist and an AI consultant. He has worked with global clients across multiple domains and helped them solve their business problems using machine learning (ML), 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 is a data scientist and ML Ops Engineer. He has worked with various global clients across multiple domains and helped them to solve their business problems extensively using advanced Machine learning (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 developing 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 also published a paper at an IEEE conference, “Deep Learning Based Approach for Range Estimation”, written in collaboration with the DRDO. He lives in Chennai with his family.

Blurb

This book will teach you how to build recommender systems with machine learning algorithms using Python. Recommender systems have become an essential part of every internet-based business today.

You'll start by learning basic concepts of recommender systems, with an overview of different types of recommender engines and how they function. Next, you will see how to build recommender systems with traditional algorithms such as market basket analysis and content- and knowledge-based recommender systems with NLP. The authors then demonstrate techniques such as collaborative filtering using matrix factorization and hybrid recommender systems that incorporate both content-based and collaborative filtering techniques. This is followed by a tutorial on building machine learning-based recommender systems using clustering and classification algorithms like K-means and random forest. The last chapters cover NLP, deep learning, and graph-based techniques to build a recommender engine. Each chapter includes data preparation, multiple ways to evaluate and optimize the recommender systems, supporting examples, and illustrations.

By the end of this book, you will understand and be able to build recommender systems with various tools and techniques with machine learning, deep learning, and graph-based algorithms.

You will:

• Understand and implement different recommender systems techniques with Python
• Employ popular methods like content- and knowledge-based, collaborative filtering, market basket analysis, and matrix factorization 
• Build hybrid recommender systems that incorporate both content-based and collaborative filtering
• Leverage machine learning, NLP, and deep learning for building recommender systems