Using the QBox platform to assess quality in data integration systems

Abstract

The data quality market is characterized by a sparse offer of tools, providing specific functions that have their own interest but are not sufficient to deal with broader user's requirements. Interoperating among these tools remains a technical challenge because of the heterogeneity of their models and access patterns. On the other side, quality analysts require more and more integration facilities that allow them to consolidate and aggregate multiple quality measures acquired from different observations, using different tools. The QBox platform aims at filling this gap by supplying (i) a generic meta-model that supports the definition of quality goals and metrics, (ii) a service-based infrastructure that allows interoperability among several quality tools and (iii) an OLAP-based quality model to support multidimensional analysis and visualization. This paper describes these three components whose usage is illustrated using a data integration scenario.