Real world data is never as perfect as we would like it to be and can often suffer from corruptions that may impact interpretations of the data, models created from data, and decisions made based on data. The problem of poor data quality stored in database -backed information systems is largely widespread in the governmental, commercial and industrial environments. It must be addressed urgently by theoretical and pragmatic approaches. These combined approaches should explore data, detect data quality problems (duplicates, errors, outliers, etc.), correct and ensure information quality in data management systems. Information quality is now one of the "hot topics" in the international academic and industrial communities involved in the management of data. Useful and efficient methods and tools that continuously inspect and monitor data quality are necessary. Very little research has been aimed at methods to support such systems.The objectives of the proposed project are to propose theoretically founded solution s for controlling in an introspective way data quality in database systems using the reflection paradigm. Abstractly, reflection refers to the capability of a system to reason about and act upon itself. EXQUALIBUR is a multi-disciplinary project, at the frontier between Database Design, Knowledge Discovery and Statistics. Its outcome is a new generation of quality-introspective data management systems. The applicant is an experienced researcher; her contributions with quality-extended query processing and quality-aware data mining are the stepping-stones of the proposed project.The host institution is AT&T Labs, which provides an excellent environment for research. Her integration into AT&T Labs will enhance the applicant's expertise in integrated management and analysis of large data volumes. Concerning the return phase, the applicant plans to apply for professorship in Europe, and to manage an independent research group on quality-aware data management.