Data quality problems beyond consistency and deduplication

Abstract

Recent work on data quality has primarily focused on data repairing algorithms for improving data consistency and record matching methods for data deduplication. This paper accentuates several other challenging issues that are essential to developing data cleaning systems, namely, error correction with performance guarantees, unification of data repairing and record matching, relative information completeness, and data currency. We provide an overview of recent advances in the study of these issues, and advocate the need for developing a logical framework for a uniform treatment of these issues. © 2013 Springer-Verlag Berlin Heidelberg.