A self-monitoring system to satisfy data quality requirements

Quality of information benefits both on line transactional processing and on line analytical processing. However, quality assurance processes are mostly human intensive and the literature provides limited support to their automation. This paper proposes a rule-based data monitoring and improvement approach as a first step towards self-management of quality of data, These rules specify when to trigger both assessment procedures and improvement actions (e.g. data cleaning), on the basis of the actions performed on the databases and specific quality requirements associated with queries performed by users. They also capture all the events occurring as a consequence of data quality problems and alert the Quality Administrator if human involvement is required. Rules are classified and formalized in the paper, The overall data quality monitoring and improvement process is explained with examples.