Risk assessment methodology for longitudinal business microdata

摘要

The paper introduces several matching approaches for simulating data attacks and thereby rating the degree of confidentiality associated with some confidential longitudinal business microdata file. From the data intruder’s perspective, the advantages and disadvantages for each of these approaches are discussed. Moreover, we analyse the possibility to combine these approaches in order to determine a global measure for the re-identification risk for some confidential microdata file. Finally, the real re-identification risk is estimated taking the example of German federal business microdata.