Following our introduction to graph structures, let's dive deeper into data-only graphs—particularly knowledge graphs—which have revolutionized the way we model and connect information.

Data Graphs: The Relationship Revolution

At their core, data graphs represent relationships between entities as nodes (vertices) and edges, where both can hold multiple data attributes. These attributes can take various forms—text, numbers, dates, and more—giving data graphs the flexibility to represent and store complex structures.

Unlike execution graphs, which define and control computational workflows, data graphs are purely about relationships and how information connects.

Core Types of Data Graphs

Data graphs come in many forms, each designed to represent specific types of real-world relationships:

• Social Networks – Graphs of people and connections (LinkedIn, Facebook)
• Road Networks – Maps with intersections (nodes) and roads (edges)
• Citation Networks – How academic papers reference each other
• Knowledge Graphs – Structured, semantic networks with meaningful relationships

While all of these capture relationships, knowledge graphs take it further by adding meaning and logic to these connections.

Knowledge Graphs: Where Meaning Meets Structure 🧠

A knowledge graph is a specialized form of a data graph where the edges contain explicit semantics—meaning they describe not just a connection, but the nature of the relationship.

🔵 What Makes a Graph a "Knowledge" Graph?

In a simple data graph, an edge simply means "connected to"—but a knowledge graph defines how and why entities are connected.

Imagine Alice isn't just a person, but specifically a doctor at a London cardiology hospital. Instead of just generic relationships, knowledge graphs embed explicit meaning into these connections.

The Triple: The Building Blocks of Knowledge

Knowledge graphs are structured using triples, which form the atomic unit of semantic information:

Subject → Predicate → Object

For example:

• Alice → is a → Doctor
• Alice → works at → London Hospital
• London Hospital → specializes in → Cardiology

These triples allow machines to reason about relationships—enabling search engines, AI assistants, and recommendation systems to "understand" data in a human-like way.

Ontologies: The Rulebook of Knowledge Graphs 📖

A knowledge graph is only as powerful as its ontology—the schema that defines how concepts relate.

For example, an ontology might state that:

• A Person can "live in" a City or "work for" a Company
• A Doctor is a specialized type of Person who "specializes in" a Field

This enables machines to infer new relationships.

For example, if Alice is a doctor at a cardiology hospital, a knowledge graph can infer that she likely has expertise in heart disease, even if not explicitly stated in the data.

Ontologies bring order and structure to knowledge graphs, ensuring that data remains consistent, interpretable, and scalable across different use cases.

Beyond Knowledge Graphs: Other Use Cases for Data Graphs 🚂

While knowledge graphs are a powerful subset of data graphs, other forms of data graphs focus on different types of relationships—including mathematical, spatial, and temporal properties.

For example, I once worked on a train visualization and rerouting system, where a train line map (from OpenStreetMap) was ingested into a graph database.

When a train broke down, the system could automatically re-calculate routes using shortest path algorithms.

In this case, the relationships between entities were not just semantic, but also included:

• Geometrical properties (distances between stations)
• Speed properties (train velocities)
• Temporal properties (estimated travel time)

These additional layers of data allowed the system to simulate and optimize route planning in real time.

This example highlights that data graphs can be much more than just knowledge representation—they can also support real-time analytics, spatial computing, and predictive modeling.

The Power of Data Graphs in a Connected World

Data graphs structure our digital world, from social media networks to financial transactions, supply chains, and AI-driven analytics.

🔹 Knowledge graphs elevate this structure by embedding meaning and logic into relationships—enabling machines to reason about complex connections the way humans do.
🔹 Other data graphs, like road networks, citation graphs, and mathematical graphs, extend beyond semantic relationships to power real-world decision-making and simulations.

Next Up: Execution Graphs

In our next post, we'll explore execution graphs, which power computational workflows—connecting data graphs with process automation and AI-driven decision-making.

Stay tuned! 🚀