Characterizing database errors

Database has become a central component of today's information systems. Nevertheless, little attempt has been made to identify the exact behavior of errors introduced to database environment. By errors, we mean defective contents spilled out of integrity enforcement established for database. Errors and their adversities, those endured rigorous integrity and constraint check built on database, must be investigated in full detail. This research is to understand the meaning of database errors, aiming to establish pro-active schemes that enable to filter out errors and prevent their propagation into information systems. Specifically, this paper attempts to classify database errors in the context of relational database, and theorizes a basic strategy for removing such errors. These issues are critical for the formalization of safety assessment and risk analysis, in light of the wide applicability of today's database applications that support real-time analysis and complex decision making. Foundations for and conceptual approach for error classification and error handling are discussed.