Objectives

The intended contributions of the QUADRIS project are as follows:

For the methodological axis: a new design method including reverse engineering of multi-source information systems and encompassing the evaluation and the control of the various facets of system quality.

For the theoretical and technical axis:

  • The definition of new strategies for measurement, for the control of various dimensions of data quality (in particular data completeness, consistency, freshness) in a context of multi-source information systems
  • A strategy for quality-driven mediation of multi-source information systems (MSIS),
  • A quality-extended metadata description and manipulation language with the associated techniques for optimization and adaptive query processing of multi-source queries driven by data quality selection and negotiation,
For the technological axis: the development of an experimental configurable middleware:

  • To detect, measure, control, explain and correct the quality of large volumes of data for any multi-source information system architecture,
  • To evaluate the global quality of a multi-source information system
  • To study the quality-driven data mediation and optimization performances of multi-source queries extended by data and system quality controls
  • To simulate, model and evaluate the cost of the proposed strategies to ensure the quality of data and the quality of the multi-source information system.
For the applicative axis: the validation of these solutions in three application domains representative for their consequent volumes of data, their complex underlying data models, and their numerous and specific data quality problems: the biomedical domain (medical records gathered by health professionals (Institute Curie)), the commercial domain (with data of EDF's Customer Relationship Management) and in the geographical domain with data on flooding and land cover (LSIS). Finally a comparative study and evaluation of the research contributions of QUADRIS compared to the current techniques of data quality management for MSIS in each application domain will be carried out and a generic framework will be proposed.
Tags:
Created by Administrator on 2007/10/01 16:27
Copyright 2004-2017 XWiki
3.0.${buildNumber}