Identifying Security Holes in OLAP Applications

Abstract

Abstract:A data warehouse system is a necessity for fundamental decisions in every enterprise. The integration of data from several internal or external sources and the functionality of modern decision support systems like OLAP tools not only provide broad access to data but also raise security problems. Security concerns are more or less the same as those of other database systems but enriched especially with access and inference control in the multidimensional model. This paper presents an integrated approach for inference and access control not on the physical but on the conceptual level. The question is not only the restriction of relations, but rather the identification and evaluation of the inference problem of hierarchies and dimensions. The possibility to restrict or perturbate data in general, is not an adequate solution. We present some specific problems of a market research company and a solution with an indicator to discover possible attacks and so be able to restrict the access by mechanisms like aggregation, restriction or perturbation.

Authorization Policies for Materialized Views

“…With the use of large systems like Data Warehouses [15] or Distributed Database Systems [3], new security issues arise [9] [13]. In our work, we focus on the problem of securing materialized views.…”