SAP SE
Anonymized data storage and retrieval

Abstract:

Systems and methods for applying k-anonymity to data from a database are provided. An initial extraction of data from a database is performed. Initial anonymized data is produced based on the extracted data and a configuration of quasi-identifiers. An actual k-anonymity level is calculated based on the initial anonymized data. The actual k-anonymity level is compared to a desired k-anonymity level. The configuration of quasi-identifiers is adjusted based on the comparison. The calculating, comparing, and adjusting are iteratively repeated until the actual k-anonymity level equals the desired k-anonymity level or the adjusted configuration of quasi-identifiers has reached a limit. Final anonymized data is produced based on the adjusted configuration of quasi-identifiers. A subsequent extraction of data from the database is performed. Subsequent anonymized data is produced based on the extracted subsequent data and the adjusted configuration of quasi-identifiers. The actual k-anonymity level is calculated based on the subsequent anonymized data.