ENTIA Systems · Knowledge framework

Entity ambiguity and identity resolution

Why a model can read a name correctly and still attach facts to the wrong business.

Definition

Entity ambiguity is the uncertainty involved in mapping a name, domain, address or identifier to the correct economic entity.

ENTIA seed proposition
“Un LLM siempre elegirá recomendar la entidad con menor Resistencia Cognitiva.”

An LLM will always choose to recommend the entity with the lowest Cognitive Resistance.

This is the seed proposition already published in ENTIA methodology. The Cognitive Resistance cluster expands the variables, evidence and operational meaning around that proposition.

Canonical source: ENTIA Methodology
What this dimension measures

What this dimension measures

Homonyms and trading names can refer to multiple legal persons.
A brand, domain and registered company may not have a one-to-one relationship.
Identifiers from different jurisdictions need normalization before they can be compared.
How ENTIA reduces the work left to the model

How ENTIA reduces the work left to the model

ENTIA resolves observations into a canonical entity and keeps the evidence used for the resolution. The machine can consume the resolved identity instead of reconstructing it from fragments.

Evidence boundary

Evidence boundary

Resolution confidence is not the same as factual truth for every attribute. Each claim still needs its own evidence and assurance.

Conceptual framework. It does not claim that lower resistance guarantees citation, ranking, recommendation or model behaviour.

Related ENTIA surfaces

Related ENTIA surfaces