Fixed-Record Feature Alternative For Variants

[orchardhillsdragon]

The fixed-record feature is used to create a single entity with variants. We tried this approach. But, we want a large set of many entities with variants randomly in name, address, SSN, MRN, date of birth, etc, and wanted to use the demographic information data set for testing EMPI matching and linking features. The fixed-record feature is not feasible to meet this requirement because it requires predefining and precompiling many entities' files.

I designed and implemented a new approach to generate demographic variants randomly, also simulating typos in name, address, SSN, MRN, date of birth, etc. I don't know if anyone has similar needs as us.