Author Description

The work described in this thesis takes a new approach to high-performance control of data quality in databases by integrally combining the ideas of integrity control subsystems and parallel data processing. Integrity control subsystems deal with requirements regarding the quality of data in databases but require high performance processing capabilities. Parallel data processing techniques are used to obtain high performance data processing both to accomodate the requirements of the integrity control subsystems and to deal with the general high demands on database management systems. This combination leads to the notion of integrity control in parallel database systems. This notion is applied in the Transaction Modification approach and prototyped on the PRISMA highly parallel prototype database machine.