KGMID and Knowledge Graph: a glossary

A KGMID is the unique identifier Google assigns to an entity in its Knowledge Graph, and it is the key every fact, attribute and knowledge panel about that entity connects back to.

The Knowledge Graph is Google's database of things rather than pages: entities, their attributes, and the relationships between them, each stored against a machine ID.

A knowledge panel is the information box Google displays when it is confident enough about an entity to present its record directly, typically at the right of desktop results.

Entity Confidence describes the degree of certainty a knowledge system holds about an entity's attributes, and it is the metric this book optimises throughout.

An entity home is the single page on a site you control that states the complete, authoritative truth about your entity, and it is the anchor every other signal resolves against.

Corroboration is agreement between independent sources on the same claim about an entity, and it is what promotes a claim into a fact the Knowledge Graph will hold.

Entity resolution is the process by which a system decides which known entity a given mention refers to, rather than treating the name as just a string of text.

Disambiguation is the deliberate separation work that lets a knowledge system tell your entity apart from anything else sharing its name.

The panel threshold is the level of confidence an entity's record must reach before Google chooses to display a knowledge panel for it.

Freebase was the collaboratively edited knowledge base Google acquired in 2010 and used as a foundation for the Knowledge Graph, and it is the origin of every KGMID beginning /m/.

Share of Answer is the proportion of AI-generated answers to a set of relevant questions in which a brand is named, used as a visibility measure in place of ranking position.

ECHO is a four-stage methodology for building visibility inside AI search, covering Entity, Corroboration, Hooks and Output, and it is run as a sequence rather than picked from.

A topical map is a planned structure of the subjects a site covers and how they relate, used to establish what an entity knows about rather than merely who it is.

Semantic distance is how far apart in meaning a query and a piece of content are, and reducing it makes content more likely to be retrieved for that query.