The ECHO Framework

Glossary term 21 of 33

Knowledge graph

A structured store of entities and the relationships between them.

What does Knowledge graph mean?

A structured store of entities and the relationships between them. Where a search engine keeps what it believes about you, as opposed to what it has merely read about you.

Quoted without alteration from the glossary of ECHO: How to Make AI Recommend Your Brand.

Which ECHO pillar does Knowledge graph belong to?

Knowledge graph belongs to Entity, pillar 1 of the ECHO framework. Does the machine know who you are?

The first ECHO pillar, and the machine's unit of understanding. A single, identifiable real-world thing with attributes and relationships, as opposed to a website, which is a collection of pages.

If the machine does not know you exist, the problem is Entity. It is built in part 2: entity, across chapter 3 and chapter 4.

  • The digital property you own and control where every core fact about your business is stated clearly, structured for machine readability, and kept current.

  • Establishing which of several possible referents a name points to.

  • The first ECHO pillar, and the machine's unit of understanding.

  • The set of web pages and data points that reference a business, with the entity home at its centre.

  • The measurable degree of certainty a search or AI system assigns to the identity of your brand as a single, disambiguated, real-world entity, built from the consistency, structure and corroboration of the facts it can find about you across the sources it trusts.

  • The specific page from which search and AI systems learn your identity, attributes and values.

  • The public-facing surface of a knowledge graph entry.

  • Name, address, phone: the core identity data for a local business.

  • A schema.org property that links an entity's canonical page to its other identity points across the web.

  • Structured data added to a page in a vocabulary machines already parse, stating facts explicitly rather than leaving them to be inferred from prose.

  • Facts published in machine-readable form, so a system can extract them without interpretation.

Nearby in the glossary

The glossary runs alphabetically. Information gain comes before Knowledge graph and Knowledge panel after it.