Incredibly useful, knowledge graphs help organizations keep track of medical research, cybersecurity threat intelligence, GDPR compliance, web user engagement, and much more. They do so by storing interlinked descriptions of entities--objects, events, situations, or abstract concepts---and encoding the underlying information. How do you create a knowledge graph? And how do you move it from theory into production?
Using hands-on examples, this practical book shows data scientists and data engineers how to build their own knowledge graphs. Authors Jesus Barrasa and Jim Webber from Neo4j illustrate common patterns for building knowledge graphs that solve many of today's pressing knowledge management problems. You'll quickly discover how these graphs become increasingly useful as you add data and augment them with algorithms and machine learning.
• Learn the organizing principles necessary to build a knowledge graph
• Explore how graph databases serve as a foundation for knowledge graphs
• Understand how to import structured and unstructured data into your graph
• Follow examples to build integration-and-search knowledge graphs
• Learn what pattern detection knowledge graphs help you accomplish
• Explore dependency knowledge graphs through examples
• Use examples of natural language knowledge graphs and chatbots
• Use graph algorithms and ML to gain insight into connected data
Author(s): Jesús Barrasa, Jim Webber
Edition: 1
Publisher: O'Reilly Media
Year: 2023
Language: English
Commentary: Publisher's PDF
Pages: 288
City: Sebastopol, CA
Tags: Data Science; Apache Spark; Apache Kafka; Graph Data Model; Cypher; GraphQL; Neo4j; Ontologies; Graph Algorithms; Metadata; Semantic Search; Knowledge Graphs; Apache Hop; Graph-Native Machine Learning; Knowledge Lake
Cover
Copyright
Table of Contents
Preface
Who This Book Is For
Conventions Used in This Book
O’Reilly Online Learning
How to Contact Us
Acknowledgments
Chapter 1. Introducing Knowledge Graphs
What Are Graphs?
The Motivation for Knowledge Graphs
Knowledge Graphs: A Definition
Summary
Chapter 2. Organizing Principles for Building Knowledge Graphs
Organizing Principles of a Knowledge Graph
Plain Old Graphs
Richer Graph Models
Knowledge Graphs Using Taxonomies for Hierarchy
Knowledge Graphs Using Ontologies for Multilevel Relationships
Which Is the Best Organizing Principle for Your Knowledge Graph?
Organizing Principles: Standards Versus Create Your Own
Creating Your Own Organizing Principle
Essential Characteristics of a Knowledge Graph
Summary
Chapter 3. Graph Databases
The Cypher Query Language
Creating Data in a Knowledge Graph
Avoiding Duplicates When Enriching a Knowledge Graph
Graph Local Queries
Graph Global Queries
Calling Functions and Procedures
Supporting Tools for Writing Knowledge Graph Queries
Neo4j Internals
Query Processing
ACID Transactions
Summary
Chapter 4. Loading Knowledge Graph Data
Loading Data with the Neo4j Data Importer
Online Bulk Data Loading with LOAD CSV
Initial Bulk Load
Summary
Chapter 5. Integrating Knowledge Graphs with Information Systems
Towards a Data Fabric
The Database Driver
Graph Federation with Composite Databases
Server-Side Procedures
Data Virtualization with Neo4j APOC
Custom Functions and Procedures
Complementary Tools and Techniques
GraphQL
Kafka Connect Plug-In
Neo4j Spark Connector
Apache Hop for ETL
Summary
Chapter 6. Enriching Knowledge Graphs with Data Science
Why Graph Algorithms?
Different Classes of Graph Algorithms
Graph Data Science Operations
Experimenting with Graph Data Science
Production Considerations
Enriching the Knowledge Graph
Summary
Chapter 7. Graph-Native Machine Learning
Machine Learning in a Nutshell
Topological Machine Learning
Graph-Native Machine Learning Pipelines
Recommending Complementary Actors
Summary
Chapter 8. Mapping Data with Metadata Knowledge Graphs
The Challenge of Distributed Data Stewardship
Datasets Connected to Data Platforms
Tasks and Data Pipelines
Data Sinks
Metadata Graph Example
Querying the Metadata Graph Model
Using Relationships to Connect Data and Metadata
Summary
Chapter 9. Identity Knowledge Graphs
Knowing Your Customer
When Does the Problem Appear?
Graph-Based Entity Resolution Step by Step
Data Preparation
Entity Matching
Build/Update a Persisted Record of Master Entities
Working with Unstructured Data
Summary
Chapter 10. Pattern Detection Knowledge Graphs
Fraud Detection
First-Party Fraud
Uncovering Fraud from Data
Fraud Rings
Innocent Bystanders
Operationalizing the Fraud Detection Knowledge Graph
Skills Matching
Organizational Knowledge Graph
Skills Knowledge Graph
Expertise Knowledge Graph
Individual Career Growth
Organizational Planning
Predicting Organizational Performance
Summary
Chapter 11. Dependency Knowledge Graphs
Dependencies as a Graph
Advanced Graph Dependency Modeling
Qualified Dependencies
Semantics of Multidependency
Impact Propagation with Cypher
Validating a Dependency Knowledge Graph
Validation 1: No Cycles
Validation 2: Aggregate Multidependencies Add Up to the Expected Total
Complex Dependency Processing
Single-Point-of-Failure Analysis
Root Cause Analysis
Summary
Chapter 12. Semantic Search and Similarity
Search over Unstructured Data
From Strings to Things: Annotating Documents with Entities
Navigating the Connections: Document Similarity for Recommendations
The Cold Start Problem
Making the Annotation Semantic with an Organizing Principle
Summary
Chapter 13. Talking to Your Knowledge Graph
Question Answering: Natural Language as a Source of Facts for a Knowledge Graph
Using Natural Language Query with a Knowledge Graph
Natural Language Generation from Knowledge Graphs
Annotating the Knowledge Graph’s Organizing Principle to Drive Natural Language Generation
Working with Lexical Databases
Graph-Based Semantic Similarity
Path Similarity
Leacock-Chodorow Similarity
Wu and Palmer Similarity
Summary
Chapter 14. From Knowledge Graphs to Knowledge Lakes
Conventional Knowledge Graph Deployments
From Knowledge Graphs to Knowledge Lakes
Looking to the Future
Index
About the Authors
Colophon
