Entity (knowledge graph)

Definition: In an AI/knowledge-graph context, an entity is a distinct, nameable thing (person, organization, product, place, concept) that systems can identify and link across sources. Entities have stable names/IDs, attributes (facts), and relationships (“X complies with Y,” “A competes with B”), enabling answer engines to disambiguate who/what a page is about and cite it safely.

Scope:

  • Includes: Canonical names and common synonyms; identifiers (e.g., Wikidata Q-IDs), attributes (type, dates, specs), and relationships; use of structured data (Schema.org) that reflects on-page facts.
  • Excludes: Vague references (“our platform,” “the solution”) without a named thing; schema dumps that don’t match visible content; purely relational database “entities” (ER models) unrelated to knowledge graphs.
  • Notes: In AEO, clear entity declaration + consistent naming improves disambiguation and source eligibility. Where possible, align with open IDs (e.g., Wikidata) and keep one canonical label per entity in your style guide.

Why it matters: Answer engines select sources they can place in a graph with low risk. If your page names its entities precisely, states how they relate, and backs claims with references, models can ground to the right node and quote you. That’s how challenger brands earn citations even against bigger domains.

See also:
Entity-rich content; Knowledge graph; Canonical names; Disambiguation; Source eligibility; Schema markup; Defined term; Internal linking (entity-aware)

References:

Synonyms: Knowledge-graph entity; Thing (in “things, not strings”); Node (graph context)