A search client calls the Identity Resolution with a search record. The Server retrieves a set of similar records from the database (known as the candidate-set).
Each candidate is scored against the search record to determine the degree of similarity. Records that score above the score-threshold defined in the Search-Definition are returned to the search client.
The process may be optimized by:
reducing the size of the candidate-set, thereby reducing the amount of scoring required, and/or
reducing the cost of scoring two records, and/or
reducing the size of the IDX to improve database cache efficiency
The following sections discuss ways in which to achieve these goals.