Data quality evaluation in data integration systems

Abstract

This thesis deals with data quality evaluation in Data Integration Systems (DIS). Specifically, we address the problems of evaluating the quality of the data conveyed to users in response to their queries and verifying if users quality expectations can be achieved. We also analyze how quality measures can be used for improving the DIS and enforcing data quality. Our approach consists in studying one quality factor at a time, analyzing its impact within a DIS, proposing techniques for its evaluation and proposing improvement actions for its enforcement. Among the quality factors that have been proposed, this thesis analyzes two of the most used ones: data freshness and data accuracy