Canonical Entity

A canonical entity is the single, authoritative, definitive version of an entity that an organization establishes as the reference identity — the agreed model of the entity that all name variants, abbreviations, distributed records, and scattered references are intended to resolve back toward. Where entity resolution is the process of connecting references to the right entity, the canonical entity is the destination that process resolves toward: the one correct, complete, authoritative version. The concept is necessary because, without a defined canonical entity, there is no single correct answer for entity resolution to reach. References to an organization exist in many forms and many places, and if there is no authoritative version designating which facts, which name, which attributes, and which relationships are definitive, systems are left to assemble a model from inconsistent inputs — and they assemble it inconsistently. Establishing a canonical entity provides the fixed reference point. It states: this is the definitive identity of the organization — its canonical name, its accurate attributes, its correct relationships, its authoritative records — and everything else should map to this. The canonical entity is closely connected to two other concepts in this lexicon. It is supported by the entity home, which is the page that publicly hosts and declares the canonical entity in structured, authoritative form. And it depends on canonical data — the single authoritative version of each underlying fact — since a canonical entity is only as coherent as the facts that compose it. For an organization, defining the canonical entity is a foundational step in entity optimization: it is the act of deciding, deliberately and definitively, what the entity is, so that entity resolution across every system has one correct answer to arrive at. Without it, entity optimization has no fixed target.