Managing with Less than Absolute Integrity

Abstract

Our key position is that for less then perfect real-world information systems the standard notions of data correctness or integrity used in research literature are inappropriate and inadequate. Hence we call upon the researchers and practitioners to pay more attention to understanding less than absolute data integrity requirements of the users and the applications, and to developing techniques to support such requirements. We will discuss this position and the need to develop criteria and techniques to support “weaker” or application specific data integrity requirements with the help of following four situations facing information systems management: data quality in multidatabase systems consistency of related (interdependent) data in multidatabase systems data integrity in process-centric environments (utilizing workflow process automation), and data integrity involving heterogeneous digital media data, such as those increasingly found in Internet and intranets. The important dimensions of data integrity that the above requirements address are: accuracy or correctness, completeness, consistency, and currentness. This position paper gives the outline of my talk.