Statistical Disclosure Control

Software tools for microdata (WP 2)

Leading partner: CBS

Participating partner: CBS

This consortium will to take over the further development of µ-ARGUS resulting from the 4th framework SDC-project. µ-ARGUS is the tool for the statistical disclosure protection of microdata.
The main reason for undertaking this extension is that the current version of µ-ARGUS is very well suited for social survey data but lacks facilities for business data. In this project we will implement several methods that will result from other workpackages (WP1.1 and WP1.2) as well as the results obtained in the SDC-project. The progress in this WP heavily depends on the work in WP1.1 and WP1.2 .
The development will be divided into 3 parts.
The testing of the software is foreseen in WP6.
The main extensions of the µ-ARGUS software are for the disclosure control of business microdata. Therefor several new techniques will be studied by the different partners in this project (see WP1.1 and WP1.2 ) and will be implemented during the course of the project. PRAM, masking techniques, micro aggregation, noise addition are techniques that will be implemented. Also disclosure risk models will be implemented.

Description of work

Task 1: Migration form Borland C++ to Visual C++, first implementation of PRAM (Post RAndoMisation), first implementation of disclosure risk.
Task 2: Implement Masking techniques (StBa), Implement the Methodological Framework (resp. UoP and IStat) They will develop software to implement the proposed methodology using S-Plus or SAS so that they can utilise statistical and graphical capabilities of these packages. In addition we will produce a full description of our algorithms. Migration plans for implementation into µ-ARGUS will be given to facilitate the implementation in µ-ARGUS.
Task 3: Implement Micro-aggregation (URV), Implement modules to take hierarchical structures into account; Implement final test results

Milestones and expected result

We aim to develop the general applicable de facto standard tool for the disclosure control of microdata. The implementation of several new techniques offers a possibility to compare the quality of these method. This might eventually lead to a coordinated approach for this topic (best practice action). The participation of several European NSI’s will guarantee the general applicability of the new version of µ-ARGUS. Besides this the results of this development can ideally be used in (TES)-courses on Statistical Disclosure Control.