During entity resolution, pipelines also complete the relationship detection process, which detects relationships between identities and entities and generates alerts for relationships of interest.
The system uses roles which are the classification of an identity that defines the focus or purpose for that identity to detect and establish relationships between entities. In the system, you define roles and then assign roles to identities by data source or as part of transforming the original data source data into Universal Message Format (UMF).
When the pipeline processes incoming identities for entity resolution and resolves the identity to an existing entity, the two records have a 0-degree relationship; that is, the incoming identity and the entity are the same. But the entity resolution process can go beyond 0-degree relationships, depending upon how the system is configured.
After the pipeline exhausts all possibilities in the entity resolution resolve phase, the relationship detection process evaluates the entities that remain on the candidate list, or those entities that did not resolve to the incoming identity, to see if a relationship exists between them. Typically, entities that are on the candidate list are linked to the incoming identity at 1-degree of separation for at least one attribute, meaning that both entities share the same attribute data values for at least one attribute, which is why the entity is on the candidate list.
After the process detects a relationship, the system compares the assigned roles between the identity and the entities to the configured role alert rules. If the system finds that the roles assigned to the identity and an entity meet the criteria for that rule, the system generates an alert, indicating it has detected a relationship of interest. The relationship can be at 0-degrees, 1-degree, or multiple degrees, depending upon how the system and the role alert rules are configured.