If you mix keys generated using multiple entities, the volume of keys increase and may negatively impact the retrieval time of candidates for matching. Databases offer options to reduce the retrieval time. For example, if you use databases to store the SSA-NAME3 keys, you may be able to partition the table that contains the SSA-NAME3 keys by entity type. You can then write the candidate retrieval queries so that the database retrieves data from each partition separately and in parallel.

Generic_Field

should be used to operate on data with high cardinality. For example, if you use SSA-NAME3 to generate keys for low cardinality data such as country code, and then try to retrieve all records in

Country=USA

. It may result in degraded performance.

Generic_Field

provides a balance between qualities of results and convenience of use. If you know the type of search data, you can use specialized algorithms, such as

Person_Name

,

Address_Part1

,

Organization_Name

to yield more accurate results.