Verfügbarkeit von personenbezogenen Daten unter dem Aspekt der Anonymisierung

Abstract

Data increasingly determines today's professional as well as private life. In recent years, personal data in particular has become increasingly relevant. While hyper-personalized advertising has faced criticism repeatedly, other areas significantly benefit from easily accessible personal information. Medical research or the development of artificial intelligence depend enormously on the availability of such data. However, new legislation is increasingly restricting its use and mandating the anonymization of personal data before it is used.

This article lists current anonymization methods and analyzes their weaknesses. The latter are critically assessed under the risk of re-identification and the accompanying loss of data quality. This paper illustrates that all techniques reduce the informative value of personal data in different ways. The article identifies three important criteria for an anonymous data set and highlights the importance of addressing them in detail. It concludes that optimal availability may require a combination of several procedures. It closes with the hypothesis that different personal attribute classes can affect the choice of an appropriate anonymization procedure.