Skip to content
BENIBIX Learn

Knowledge Graph

Also known as Knowledge Graphs, Entity Graph

Facts stored as things and the links between them, like a map you can follow.

Think of it like

It is like a family tree. Dots are people, and lines say how they are linked.

Example

A graph holds "Emma works at Acme" and "Acme is in London". Asked where Emma works, the system follows two links and answers London.

Why it matters

Some answers need two or three facts joined together. A graph keeps those links ready to follow.

Definition

A knowledge graph stores facts as things and the links between them. Each thing, like a person, a city or a company, is a node. Each link is a labelled line, like "works at" or "born in". You can follow the lines to answer questions that hop across facts. Search engines use them for the fact boxes beside results. AI systems use them to find linked facts that plain text search misses.
Each dot is a thing and each line is a named link between two things.

Common mistake

Thinking it builds itself. Someone, or a model, must pull out the things and links, and keep them up to date.

Key takeaways

  1. 1
    Things are nodes. People, places, products.
  2. 2
    Links have names. Works at, born in, part of.
  3. 3
    Follow the lines. Hop across facts to answer.

Compare