book

Scaling Graph Learning for the Enterprise

by Ahmed Menshawy, Sameh Mohamed, Maraim Rizk Masoud

August 2025

Intermediate to advanced

368 pages

11h 15m

English

O'Reilly Media, Inc.

Read now

Unlock full access

Includes

Quizzes

Preface
What Is Graph Learning for the Enterprise?Why Does Graph Learning Matter for the Enterprise?Who Is This Book For?Overview of the ChaptersConventions Used in This BookUsing Code ExamplesO’Reilly Online LearningHow to Contact UsAcknowledgements
1. Introduction to Graphs
The Power of Enterprise Graph Learning and Inference at ScaleA Bird’s-Eye View: Navigating the Book’s ChaptersGraphs and Graph LearningWhat Is a Graph?Graph Data RepresentationGraph LearningScalable Graph Learning: Addressing the RequirementsAdvantages of Scalable Graph Learning in EnterpriseLarge-Scale Graphs in Real-World Enterprises: Use CasesTravel-Time Predictions on Google MapsDrug Development: HalicinFraud DetectionThe Evolution of Graphs and Graph Learning: From Early Beginnings to Modern ApplicationsEra 1: The Foundation of Graph Theory and Algorithms (1736-1970)Era 2: More Advancement in Graph Algorithms and Technologies (1970-1999)Era 3: Emergence of Graph Databases and Graph Query Languages (2000-2006)Era 4: Graph Analytics and Traditional Machine Learning (2007-2011)Era 5: Rise of Graph Neural Networks (2012-2018)Era 6: Scalability, Robustness, and Enterprise Applications (2019-Present)Challenges of Enterprise-Ready Graph Learning SystemsData Harmonization ChallengesComputationally Intensive WorkloadsDynamic Evolving GraphsActive Monitoring and Drift DetectionReal-Time InferenceSummary
2. The Graph Machine Learning Pipeline
The Graph Data PipelineDefinition of Graph Data and Graph Data LevelsGraph Data Sourcing and UnderstandingGraph Data PreparationThe Graph Training and Inference PipelinesGML Training Pipeline OverviewGML Inference Pipeline OverviewSummary
3. Traditional Machine Learning for Graphs
Approaches to Graph Machine LearningTraditional Graph-Based Machine LearningNontraditional Graph-Based Machine LearningRepresenting Graphs for Traditional MLGraph RepresentationRepresenting Amazon Copurchasing Networks as GraphsNavigating Graph Tasks in the Amazon Copurchasing DatasetGraph Feature EngineeringImportance and ChallengesTypes of Graph FeaturesHands-on: Extracting Features for the Amazon Copurchasing GraphGraph Features in ML ModelingTask and Techniques OverviewPredicting High-Rated Products with a Prediction ModelFeature Learning with Node EmbeddingsRandom Walk AlgorithmAmazon Copurchasing Dataset and Node EmbeddingsSummary
4. PyGraf: End-to-End Graph Learning and Serving
Graph Libraries OverviewChallenges of Open Source Graph Libraries: PyGraf OpportunitiesPyGraf: A Solution for Streamlined Graph Learning and ServingIntroduction to PyGrafPyGraf Key FeaturesPyGraf Purposes: Empowering Dynamic EnvironmentsArchitecture and Core CapabilitiesCore Components LayerAdaptation and Integration LayerBest Practices LayerIn-Depth Exploration of Core Library ComponentsData ComponentTraining ComponentServing ComponentPrivacy Preserving ComponentEnd-to-End Example Using PyGraf: Amazon Copurchasing DatasetPreprocessing and TransformationModel TrainingEvaluation and Model SelectionDeployment and MonitoringSummary
5. Graph Neural Networks
Introduction to Graph Neural NetworksThe Significance of GNNs in Graph LearningOverview of GNN ApplicationsFoundations of GNNsGraph Convolutional NetworksTransition from Traditional Graph Learning to GCNsHow GCNs Simplify Feature EngineeringHands-on E2E Example Using PyG and the PyGraf InterfacePyG: Karate Club ExampleCora Node ClassificationPyGraf API InterfaceLimitations of GCN-Based ArchitectureSummary
6. Advanced Techniques in Graph Learning
Different Types of GraphsHomogeneous GraphsHeterogeneous GraphsTemporal GraphsGraph Embedding ModelsHow Do Knowledge Graph Embeddings Work?Training Knowledge Graph Embedding ModelsEmbedding Interaction MethodsTraining ObjectivesStrengths of Graph Embedding ModelsExample: Learning on Freebase DatasetAttention on GraphsFeature VectorsAttention MechanismMultihead AttentionExample: Citation NetworkCiteSeer Use Case ExampleData Preprocessing and VisualizationModel Training Using PyGModel TestingEmbeddings VisualizationSummary
7. Scalable Graph Neural Networks
Challenges in Scaling Graph Learning ModelsComputational ChallengesMemory ConstraintsData Complexity and SizeMini-Batching in Graph Neural NetworksDefinition and Importance of Mini-BatchingTechniques for Effective Mini-BatchingMemory-Efficient Training TechniquesGradient CheckpointingSubgraph Sampling MethodsLayer-Wise Relevance PropagationDistributed Data and Compute StrategiesDistributed Execution StrategiesGraph Partitioning StrategiesDistributed Graph Learning ToolsDistributed Training with PyGPrepare and Partition the Graph DataManual Execution of Distributed TrainingKey Components of Distributed Training with PyGAdvanced Architectures for Scaling Graph Neural NetworksSparsity ExploitationApproximation TechniquesUse of External MemorySummary
8. Enterprise Applications of Graphs
Customer and Market InsightsCustomer SegmentationSocial Network AnalysisRecommendation EnginesOperations and Supply Chain ManagementNetwork Optimization in Supply ChainsInventory ManagementRisk Management and Supply Chain ResilienceReal-World ApplicationsSecurity and Risk ManagementThreat Detection and AnalysisIdentity and Access ManagementFraud DetectionCybersecurity Incident ResponseRisk Management and ComplianceHealthcare and Life SciencesPatient Journey Mapping and PersonalizationDrug Discovery and DevelopmentClinical Trial OptimizationGenomics and Personalized MedicinePopular Life Science ProductsSummary
9. Privacy-Preserving Graph Learning
The Importance of Privacy in Graph LearningEnterprise Examples and ApplicationsOverview of Privacy-Preserving TechniquesPrivacy Threats in Graph LearningCase Studies of Privacy BreachesPrivacy-Preserving Techniques for Graph DataGraph Data Anonymizationk-Anonymity and t-Closeness in GraphsGraph Modification TechniquesEdge Differential PrivacyNode Differential PrivacySynthetic Graph Data GenerationPrivacy-Preserving Graph ComputationSMPC for GraphsHomomorphic Encryption in Graph LearningDifferentially Private Learning AlgorithmsFederated Graph LearningPrinciples and ArchitectureFederated Graph Neural NetworksTraining and Communication ProtocolsPrivacy Enhancements in Federated Graph LearningApplications of Privacy-Preserving Graph LearningPrivacy-Preserving Techniques in Social Network AnalysisUse CasesCase Study: Applying FedGraphNN to a Recommender System Using Epinions DataDatasetImportance of Applying Federated LearningData ProcessingFederated Training LogicRunning the Federated PipelineSummary

10. Graph Inference and Deployment Strategies
Deployment StrategiesCanary Deployments: Testing the WatersBlue–Green Deployments: Smooth TransitionsShadow Deployments: Flying Under the RadarEdge Deployments: Taking It to the EdgeA/B Testing Deployments: Controlled ExperimentsProgressive Rollouts: Scaling Up GraduallyContainerized DeploymentsHardware ConsiderationsInference RuntimesInference Frameworks and LibrariesPrecomputations and CachingOnline Versus Offline InferenceModel Optimization TechniquesQuantizationPruningKnowledge DistillationScaling Inference for Large GraphsDistributed Inference SystemsGraph Partitioning for Scalable InferenceIncremental InferenceSummary
11. Monitoring and Feedback Loops
Challenges in Monitoring Graph ModelsGraph-Specific MetricsDynamic GraphsScale and ComplexitySystem and Deployment IssuesDesigning a Monitoring FrameworkDealing with Ever-Changing Connections (Graph Topology)Checking the “Meaning” (Evaluating Embeddings)The Balancing Act (System Performance Versus Graph Tasks)Data Collection MechanismsMonitoring Deployed Graph ModelsThresholds for AlertsVisualization ToolsFeedback Loops in Graph SystemsUser Feedback: Learning from InteractionSystem Feedback: Optimizing Model PerformanceData Feedback: Adapting to a Changing WorldSynergy Between Feedback LoopsClosed-Loop Systems: Automating AdaptationAdaptive Retraining PipelinesNotification SystemsSummary
12. Future Trends: Graph Learning and LLMs
Introduction to Graph-Enhanced LLMsLLMs and Their Transformative ImpactThe Need for External Memory: The RAG SetupBenefits of Graph-Enhanced RAG SetupRetrieval Augmentation with Graph IntegrationChallenges in Traditional Retrieval-Augmented GenerationGraph IntegrationGraphRAG Methodology and PipelineIndexing ProcessQuerying ProcessEnhanced Capabilities and Whole-Dataset ReasoningImplementation ConsiderationsBaseline RAG Versus GraphRAGRAG with Knowledge Graphs for Customer Service Question AnsweringSummary
Index
About the Authors

Content preview from Scaling Graph Learning for the Enterprise

Chapter 4. PyGraf: End-to-End Graph Learning and Serving

Now that you’ve grasped the graph pipeline and its traditional ML approaches, this chapter introduces our open source graph machine-learning library, PyGraf. It serves as our framework for constructing robust graph learning ecosystems at an enterprise level.

Throughout this chapter, we’ll walk you through the library’s architecture and its core modules. We’ll also showcase the library’s capabilities by reworking an example presented in Chapter 3 using this library.

By leveraging this library, you, as a reader, will not only solidify your understanding of concepts from earlier chapters, but also learn how to practically apply them in common tasks. You will also be able to experience graph representation, and explore different ways to utilize this process. Additionally, the library encompasses an entire GML pipeline, featuring a range of algorithms, metrics, and connectors. This facilitates the seamless creation of custom pipelines with precision and ease. With these resources readily available, you will have comprehensive guidance to navigate and follow along in the upcoming chapters.

Let’s delve into this chapter by exploring the library and uncovering its purpose.

Graph Libraries Overview

The emergence of graph learning and serving technologies has opened new frontiers in data analytics, particularly for enterprises dealing with large-scale, complex datasets. With this open source and prosperity GML libraries have emerged ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781098146054Errata Page

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills

Scaling Graph Learning for the Enterprise

by Ahmed Menshawy, Sameh Mohamed, Maraim Rizk Masoud

Chapter 4. PyGraf: End-to-End Graph Learning and Serving

Graph Libraries Overview

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.