Automated data consistency checking using SBVR: Case study: Academic data in a University

摘要

Data are the most important asset of an organization. Even though the application or business rules change, consistent data stored in the database must be kept. This research is intended to propose a framework for checking the consistency of an existing database to the business rules in which the data must comply. The research focuses on the situation where the documents of database and the application are obsolete. The rules are extracted from the associated documents (e.g. the decrees) before they are finally expressed in SBVR. Those rules are used for data checker SQL script generation. The automation process is possible if the database has a naming convention and the organization has a standardized vocabulary. The framework has been tested using academic data, where the database has no documentation, and the application is still operating until today. A business rule grammar and SQL query templates are proposed. Data consistency checker tool based on the academic rules expressed in SBVR is developed. The research conducted found data that do not comply with the rules. The contribution of this research is a framework and a tool for data consistency checking.