A Comparative Study of Microaggregation Methods

Microaggregation is a statistical disclosure control technique for microdata. Raw microdata (i. e. individual records) are grouped into small aggregates prior to publication. Each aggregate should contain at least k records to prevent disclosure of individual information. Fixedsize microaggregation consists of taking fixed-size microaggregates (size k). Data-oriented microaggregation (with variable group size) was introduced recently. Regardless of the group size, microaggregates on a multidimensional data set can be formed using univariate techniques on projected data or using multivariate techniques. This paper presents the first method for multivariate fixed-size microaggregation. In addition, a real data set is used to compare the information loss and output data quality of fixed-size vs. data-oriented, and univariate vs. multivariate microaggregation.