Abstract: Generating and providing match justifications in association with record matching results for improved record matching is provided. An entity resolution system generates a set of deterministic rules based on patterns of attribute comparison outcomes of known matched and unmatched records. Each rule includes matching conditions and an action instruction including a message of a justification for a match/non-match. The system receives a request to determine a match between two records. A matching engine compares various attributes of the records using probabilistic matching technologies to determine a match outcome. A justification engine compares attribute comparison outcomes to the stored rules. When the matching conditions of a rule are satisfied, the rule is activated and a human-readable justification is linked to the match output and provided in a match response. The justification provides insights into the match outcome, and improves the speed and accuracy of a verification process of the match output.

Abstract: A request for a most reliable demographic fields data for a selected demographic field associated with an entity may be received at an entity matching system. A plurality of records associated with the entity may be received at the entity matching system. Each of the plurality of records may include at least one demographic field. A demographic field corresponding to the selected demographic field may be identified from the at least one demographic field for each of the plurality of records. A reliability score may be determined for the corresponding demographic field for each of the plurality of records. The corresponding demographic field with the highest reliability score may be determined as the most reliable demographic fields data.

Abstract: Generating and providing match justifications in association with record matching results for improved record matching is provided. An entity resolution system generates a set of deterministic rules based on patterns of attribute comparison outcomes of known matched and unmatched records. Each rule includes matching conditions and an action instruction including a message of a justification for a match/non-match. The system receives a request to determine a match between two records. A matching engine compares various attributes of the records using probabilistic matching technologies to determine a match outcome. A justification engine compares attribute comparison outcomes to the stored rules. When the matching conditions of a rule are satisfied, the rule is activated and a human-readable justification is linked to the match output and provided in a match response. The justification provides insights into the match outcome, and improves the speed and accuracy of a verification process of the match output.

Abstract: A request for a most reliable demographic fields data for a selected demographic field associated with an entity may be received at an entity matching system. A plurality of records associated with the entity may be received at the entity matching system. Each of the plurality of records may include at least one demographic field. A demographic field corresponding to the selected demographic field may be identified from the at least one demographic field for each of the plurality of records. A reliability score may be determined for the corresponding demographic field for each of the plurality of records. The corresponding demographic field with the highest reliability score may be determined as the most reliable demographic fields data.