Modeling Data as Graphs Learning Plan

MDG1: Overview of Graphs from a Data Modeling Perspective

• Which Graph Models?
• Two Major Categories: Property Graph vs. RDF
• Semantic Web – RDF: Resource Description Framework
• OWL: Web Ontology Language - W3C Standards
  • RDF Stores are Graph DBMS’s
  • Property Graph – Cypher Example
  • What About Knowledge Graphs?
  • Several Communities
  • Property Graph Data Model
  • The Basics of Property Graphs
• “Everything Looks Like a Graph”
  • Cognition And Perception
    • Conceptual Spaces
    • Spatial Cognition
    • Spatial Thinking is the Foundation of Abstract Thought
  • Make Complex Financials
    Simple
  • The Visual Syntax
  • IMDB-Inspired Data Model for Movies
  • Primary Takeaway
    • Exercise
• Modeling Approaches
  • Full-Scale Data Architecture
    • Two-by-Two Dimensions: Manage the Concerns, which are Competing
  • The three-Layered Architecture
  • Agile
  • Control vs. Flexibility
• “Super Data Models” (Schema First)
  • Five Steps
• From Business Concepts to Nodes and Relationships
  • Overview of a Car Dealership
  • A Simple Concept Model
  • Structure and Meaning Expressed as A Property Graph
  • Property Graph Databases
  • Timetree Library
  • Email Concepts
  • Do Graph Databases Really Do Many-to-Many?
    • How to Deal with Many-to-Many
  • What Goes into a Good Relationship?
• “Fast Track Data Models” (Schema Last)
  • Three Steps
• Demo Scripts Example and Practice – Load and Analyze Data
  • Where is the Target Data Model? In the Schema?
• Physical Considerations
• Summary: Why Is Graph Data Modeling Different?
  • Strongholds of Property Graphs
  • Differences Between Relational Normalization and Graph Data Modeling

MDG2: Property Graph Data Modeling Compared to Classic Data Modeling

• The History of DBMS Technologies
  • The Pioneers of the DBMS Trail
  • The Relational Empire And Then NoSQL
  • Peter Chen – Entities, Attributes, Relationships
    1976
• What Is Different in Graph Data Modeling? (Recap)
  • Structure and Meaning Expressed as A Property Graph
  • Why is Graph Data Modeling Different?
• Visualization
  • The Basics of Graphing Data
  • A Concept Model (Map): ”A Data-Driven User Story”
  • Concept Models Come from Educational Psychology
  • Property Graph Representation
• Highly Connected Data
  • Empirical Data About Relationships in Databases
• Normalization
  • The Relational Model: Relations or Relationships?
  • The Relational Model: Making Sense by Exploring Relationships?
  • Draw a Naive Concept Map
    • Visualize Implicit Relationships
    • Name the Relationships
    • Resolve the Functional Dependency of City
    • Visualize the Properties and Finalize the Concept Design
  • The List of Fields to be Modeled
  • Five Steps in the Process
  • Two types of Dependencies
  • Isolate independent Multiple Relationships?
  • Isolate Semantically Related Multiple Relationships?
  • The Contemporary Style of Normalization
  • A Naive Concept Map of Suppliers (2nd Normal Form)
  • A Comprehensive Concept Map
  • Property Graph Representation
  • Denormalization Creates More Data
• Many to Many
  • Is It Really Many-to-Many?
• Think in Paths
  • Cypher Query Structure
  • Simple Repository Model
  • Query to Build The Lineage Solution
• Labels (and Types)
  • About Labels
  • Modeling with Labels
• The Role of the Schema (if any)
  • Design Thinking
  • The Process
  • Schema Evolution
  • Constraints
  • Schema First
  • Schema Last
• New Opportunities for the Data Modeler
  • Identity
  • Uniqueness
  • Best Practice Data Modeling and Graph Data Modeling
  • What Goes into a Good Relationship?
  • Things to Remember
• Summary
  • List of Concepts
  • The Connected Data Model
  • The Structured Data Model
  • The Governable Data Model
  • Summary of the Differences Between Relational Normalization and Graph Data Modeling

MDG3: Cookbook for Modeling Business Data as Property Graphs 

• “Everything Looks Like a Graph” (recap)
• Property Graph Data Model
• The Basics of Graphing Data
• The Basics of Property Graphs
• ”Lifting the Data Model off the Whiteboard…”
• The Visual Syntax
• Primary Takeaway
• Modeling Approaches
• Full-Scale Data Architecture
  • Two-by-Two Dimensions: Manage the Concerns, which are Competing
• General and Business-Level Concerns
• Solution and Physical-Level Concerns
• The Three-Layered Architecture
• Agile
• Control vs. Flexibility
• The Concern-driven To-Do List

• Build Your Own Repository
• A Simple Repository to Use
• The Repository (Maintained by the Scripts)
• Two General Meta-Properties
• The Concept Model Area
• The Business-oriented Concept Parts of the Repository
• Parts of a Script for Creating Concept Model Parts
  • A Sample Concept Model
• Solution Models
  • Parts of A Sample Script to Load a Solution Model
  • Sample Solution Model
• Physical Models
  • Sample Script for Creating Parts of a Physical Model
  • Sample Physical Model
• Lineage in the Repository
• Metadata in the Repository

• Super Data Models
• Five Steps
• How To Get a Concept Model To Work
  • Your Options
• Overview of a Car Dealership
  • Exercise
• A Simple Concept Model, Self-References?
• Which are the “Universal Component Parts” of Data Models?
• “The Atoms and Molecules” of Data Models
• ”Molecular Structure” of the Issues to Address
• The Concern-driven To-Do List

• Recycling Concept Maps into Graph Data Models
• Brainstorming Using CmapTools and a Data Projector
• The CXL File
• The Structure of the XML Document
• The Graph Representing the Concept Map
• The Extracted Concept Model
  • Verify Your Concept Model

• Build a Solution Model and a Physical Model
• Steps to Follow During this Exercise
• The Result is This Lineage Diagram
• Extend (or Reduce) the Solution Data Model
• Generate a Physical Data Model
• Two Scripts to Generate Load Templates
• Other Things You Could Consider
• Property Graph Databases - The Ubiquitous Pointer!
• Timetree Library
• Email Concepts
• What Goes into a good Relationship?
• Many-to-Many
• Concerns
• Summary
  • List of Concepts
  • The Connected Data Model
  • The Structured Data Model
  • The Governable Data Model
  • Summary of the Differences Between Relational Normalization and Graph Data Modeling

MDG4: Agile Graph Data Model Evolution

• “Everything Looks Like a Graph” (recap)
• Property Graph Data Model
• The Basics of Graphing Data
• The Basics of Property Graphs
• ”Lifting the Data Model off the Whiteboard…”
• A Concept Model
• Primary Takeaway
• Modeling Approaches (recap)
• Full-Scale Data Architecture
• Two-by-Two Dimensions: Manage the Concerns, which are Competing
• Comparing Concerns
• Agile
• The To-Do List
• Overview of the Loading Options for Graphs
• Ways to Load Data into a Graph Database
• “Hands-free” Loading from a CSV-file
• Neo4j ETL Tool
• LOAD From CSV, JSON, etc.
• APOC Library
• Two Major Challenges
• Business Semantics
• The Human Factors
• Overview of a Car Dealership
• ”Molecular Structure” of the Issues to Address
• What to Check in Your Graph Data Model
• Structure and Meaning Expressed as A Property Graph
• Bottom-Up vs. Schema LESS
• Profiling Your Graph Data (Example Scripts Included)
• Inferring the Schema From The Data
• How Many Node Types and Relationships?
  • Count Nodes
  • Which Properties Exist in Nodes
• Evaluate the Uniqueness of a Property within a Label
• Which Relationship Types?
  • Which Properties on Relationships?
  • Count Relationship Types Occurrences
• Finding M:M Cardinalities across Relationships
• Graph Structures
• The Power of Dependencies
• Identity
• Uniqueness
• Best Practices for Graph Data Modeling (a Checklist)
• Data and Relationship Names
• Identity and Uniqueness
• Keys
• Simplified Email Property Graph
• State Changes
• Versions
• Proper Housekeeping
• Relationships and Missing References
• The Right Level of Detail
• Trinary Relationships or Properties on Relationships?
• One-to-one, Self References
• Relationship Quality
• Data Names
• Overview of Graph Refactoring
• Part of the APOC Library
• Summary
  • List of Concepts
  • The Connected Data Model
  • The Structured Data Model
  • The Governable Data Model
  • Summary of the Differences Between Relational Normalization and Graph Data Modeling