Methodological Guidelines and Adaptive Statistical Data Validation to Build Effective Data Warehouses

Over time, data integration involving data warehouses is becoming more difficult to develop and to manage due to the growing heterogeneity of data sources. Despite the significant advances in research and technologies, many integration projects are still too slow to generate pragmatic results and are often abandoned before that. The objective of this work is the specification of a developing strategy to build effective data warehouse integration systems with progressive increase of data quality. For the achievement of our goal, we propose some methodological guidelines for the construction of evolutionary integration systems and the development of a framework for automated statistical validation that will facilitate the enhancement of system’s data quality.