Relationship Modeling

🏘️ NoSQL Relationship Modeling: A Story of Connections

Imagine you’re building a neighborhood. Houses need to connect to each other, to people, to streets. Some connections are simple—one person lives in one house. Others are complex—many friends visiting many houses. Let’s explore how data in NoSQL databases connect, just like neighbors in a town!

🎯 What You’ll Learn

• One-to-One Relationships (One thing connects to exactly one other thing)
• One-to-Many Relationships (One thing connects to many things)
• Many-to-Many Relationships (Many things connect to many things)
• Cardinality (Counting how many connections exist)

🏠 The Universal Analogy: The Neighborhood

Think of your database as a neighborhood.

• Documents are like houses
• Relationships are like pathways between houses
• Cardinality tells us how many pathways connect things

Let’s walk through this neighborhood together!

1️⃣ One-to-One Relationships

🤝 What Is It?

One thing connects to exactly ONE other thing.

Like having ONE best friend. You have one. They have one. That’s it!

🏠 Neighborhood Example

One House → One Mailbox

Every house has exactly one mailbox. Every mailbox belongs to exactly one house.

graph TD
    A["🏠 House"] --> B["📬 Mailbox"]

💻 NoSQL Example

Method 1: Embed Inside (Keep Together)

Store the related data INSIDE the main document:

{
  "_id": "user_123",
  "name": "Emma",
  "passport": {
    "number": "AB123456",
    "country": "USA",
    "expires": "2030-01-15"
  }
}

Why? Emma has ONE passport. Keep it together!

Method 2: Reference (Keep Separate)

Store a link to another document:

{
  "_id": "user_123",
  "name": "Emma",
  "passport_id": "passport_789"
}

Separate passport document:

{
  "_id": "passport_789",
  "number": "AB123456",
  "country": "USA",
  "expires": "2030-01-15"
}

✨ When to Use Each?

🎨 Real-World Examples

• User → Profile Settings (embed)
• Person → Birth Certificate (reference)
• Car → License Plate (embed)
• Employee → Tax Record (reference)

2️⃣ One-to-Many Relationships

📚 What Is It?

ONE thing connects to MANY things.

Like a parent with children. One parent → many kids!

🏠 Neighborhood Example

One House → Many Windows

A house has many windows. Each window belongs to one house.

graph TD
    A["🏠 House"] --> B["🪟 Window 1"]
    A --> C["🪟 Window 2"]
    A --> D["🪟 Window 3"]
    A --> E["🪟 Window 4"]

💻 NoSQL Example

Method 1: Embed Array (Many Inside One)

Put all related items inside the parent:

{
  "_id": "blog_post_1",
  "title": "My Day at the Zoo",
  "comments": [
    {
      "user": "Alice",
      "text": "So fun!"
    },
    {
      "user": "Bob",
      "text": "Love the photos!"
    },
    {
      "user": "Carol",
      "text": "Which zoo?"
    }
  ]
}

Method 2: Reference Array (Links to Many)

Store IDs pointing to separate documents:

{
  "_id": "author_1",
  "name": "J.K. Rowling",
  "book_ids": [
    "book_1",
    "book_2",
    "book_3"
  ]
}

Method 3: Parent Reference (Child Points Back)

Each child stores the parent’s ID:

{
  "_id": "order_item_1",
  "product": "Pizza",
  "order_id": "order_100"
}

{
  "_id": "order_item_2",
  "product": "Soda",
  "order_id": "order_100"
}

✨ Choosing the Right Pattern

🎨 Real-World Examples

• Author → Books (one author writes many books)
• Customer → Orders (one customer makes many orders)
• Category → Products (one category has many products)
• Teacher → Students (one teacher teaches many students)

3️⃣ Many-to-Many Relationships

🕸️ What Is It?

MANY things connect to MANY other things.

Like students and classes. One student takes MANY classes. One class has MANY students!

🏠 Neighborhood Example

People ↔ Clubs

Many people join many clubs. Many clubs have many members.

graph TD
    A["👤 Alice"] --> D["📖 Book Club"]
    A --> E["🎮 Gaming Club"]
    B["👤 Bob"] --> D
    B --> F["⚽ Sports Club"]
    C["👤 Carol"] --> E
    C --> F

💻 NoSQL Example

Method 1: Two-Way References

Store arrays on BOTH sides:

Student document:

{
  "_id": "student_1",
  "name": "Emma",
  "class_ids": ["math_101", "art_202"]
}

Class document:

{
  "_id": "math_101",
  "name": "Basic Math",
  "student_ids": ["student_1", "student_2"]
}

Method 2: Join Collection (Bridge Table)

Create a THIRD collection to link them:

{
  "_id": "enrollment_1",
  "student_id": "student_1",
  "class_id": "math_101",
  "enrolled_date": "2024-01-15",
  "grade": "A"
}

Why use a join collection?

• Store extra info about the relationship (grade, date)
• Handle unlimited connections
• Avoid array size limits

Method 3: Embed on One Side

When one side has fewer items:

{
  "_id": "book_1",
  "title": "Database Design",
  "authors": [
    {"id": "author_1", "name": "Alice"},
    {"id": "author_2", "name": "Bob"}
  ]
}

✨ Pattern Comparison

🎨 Real-World Examples

• Students ↔ Classes (join collection with grades)
• Products ↔ Tags (two-way references)
• Users ↔ Roles (embed roles in users)
• Doctors ↔ Patients (join collection with appointments)

4️⃣ Cardinality

🔢 What Is It?

Cardinality = How many connections exist on each side.

It’s like counting: “How many?” on each end of the relationship.

🎯 The Four Types

graph TD
    A["Cardinality Types"] --> B["1:1 One-to-One"]
    A --> C["1:N One-to-Many"]
    A --> D["N:1 Many-to-One"]
    A --> E["N:M Many-to-Many"]

📊 Visual Guide

💡 Why Cardinality Matters

It helps you CHOOSE the right pattern!

Decision Tree

graph TD
    A[What's the cardinality?] --> B{1:1?}
    B -->|Yes| C["Embed or Reference"]
    B -->|No| D{1:N or N:1?}
    D -->|Yes| E{How many N?}
    E -->|Few <100| F["Embed Array"]
    E -->|Many| G["Reference or Parent Ref"]
    D -->|No| H["N:M → Join Collection"]

🔍 Estimating Cardinality

Ask yourself:

• Low (1-10): Embed easily
• Medium (10-100): Embed with caution
• High (100-1000): Use references
• Unbounded (1000+): Always use references

Example Analysis

Blog Post → Comments

“How many comments per post?”

• New blog: ~10 comments → Embed ✅
• Popular blog: ~1000 comments → Reference ✅

// Low cardinality (embed)
{
  "post": "Hello World",
  "comments": ["Great!", "Nice!"]
}

// High cardinality (reference)
{
  "post": "Viral Video",
  "comment_count": 50000
}
// Comments in separate collection

🎯 Quick Summary

🏆 You Did It!

You now understand:

✅ One-to-One: One thing → One thing (Person → Passport)

✅ One-to-Many: One thing → Many things (Author → Books)

✅ Many-to-Many: Many things ↔ Many things (Students ↔ Classes)

✅ Cardinality: Counting connections to choose the right pattern

Remember our neighborhood:

• Houses (documents) connect via pathways (relationships)
• Count the pathways (cardinality) to build the best neighborhood!

Now go build amazing connected data! 🚀