Knowledge Graph is a structured way of representing information as a network of entities (things) and the relationships between them. It lets machines store and reason about facts and how they connect.
What it means in plain English
A knowledge graph organises facts as a web: nodes are entities (people, places, products), and links are the relationships between them (“was born in,” “is made by”). This connected structure lets systems answer questions that require joining facts together, and understand context rather than isolated data points. Search engines and voice assistants use knowledge graphs to give direct, factual answers.
It captures not just data, but how pieces of data relate.
A simple example
When you search a public figure’s name and get a box of facts — their birthplace, occupation, and related people — that comes from a knowledge graph connecting those entities and relationships behind the scenes.
Why it matters
Knowledge graphs power factual search results, recommendation systems, and question-answering. By capturing relationships between facts, they help machines reason about the world more like humans do, and they are increasingly combined with language models to improve accuracy.
Related terms
- Named Entity Recognition — often used to populate knowledge graphs.
- Semantic Search — enhanced by knowledge graphs.
- Grounding — using knowledge graphs to keep AI factual.
Frequently asked questions
What is a knowledge graph?
It is a structured network of entities (people, places, things) and the relationships between them, letting machines represent and reason over connected facts.
How are knowledge graphs used with AI?
They power search, recommendations, and question answering, and can ground AI systems in structured, verifiable facts to improve accuracy.
