Skip to main content

🗓️ 22012025 2344
📎

data_models

Hierarchical Model​

Represents data as one big tree.

  • Use Case: Organizational charts, file systems.
  • Strengths:
    • Simple to understand and implement
    • Effective for strictly hierarchical data
  • Weaknesses:
    • Poor at handling many-to-many relationships
    • Changes to hierarchy require schema changes.
  • Example: IBM IMS, Windows Registry.

Relational Model​

Based on a structured schema of tables, designed to address the limitations of the hierarchical model

  • Use Case: E-commerce platforms, banking systems.
  • Strengths:
    • Flexible and widely used.
    • Strong support for data integrity and constraints.
  • Weaknesses:
    • Struggles with unstructured data.
    • Joins can be a bottleneck in large datasets.
  • Example: PostgreSQL, MySQL, Oracle.

Non-Relational Models​

Document Model​

  • Description: Data is stored as self-contained documents (e.g., JSON, BSON).
  • Use Case: Product catalogs, content management systems.
  • Strengths:
    • Great for unstructured/semi-structured data.
    • Scales well horizontally.
  • Weaknesses:
    • Relationships between documents are harder to enforce.
  • Example: MongoDB, Couchbase.

Graph Model​

Designed for highly connected data with complex relationships (nodes and edges)

  • Use Case: Social networks, recommendation systems, fraud detection.
  • Strengths:
    • Efficient for traversing relationships.
    • Intuitive for graph-like data.
  • Weaknesses:
    • Less mature and standardized than relational databases.
  • Example: Neo4j, JanusGraph.

Transition to Non-Relational Models​

  • Why Non-Relational Models Emerged:
    • Big data and distributed systems require better horizontal scalability.
    • Increasing need to handle semi-structured and unstructured data.

References

  • DDIA Chapter 2