Engineering privacy requirements in business intelligence applications

In this paper we discuss the problem of engineering privacy requirements for business intelligence applications, i.e., of eliciting, modeling, testing, and auditing privacy requirements imposed by the source data owner on the business intelligence applications that use these data to compute reports for analysts. We describe the peculiar challenges of this problem, propose and evaluate different solutions for eliciting and modeling such requirements, and make the case in particular for what we experienced as being the most promising and realistic approach: eliciting and modeling privacy requirements on the reports themselves, rather than on the source or as part of the data warehouse. © 2008 Springer-Verlag Berlin Heidelberg.