Business information modeling: A methodology for data-intensive projects, data science and big data governance

Abstract

This paper discusses an integrated methodology to structure and formalize business requirements in large dataintensive projects, e.g. data warehouses implementations, turning them into precise and unambiguous data definitions suitable to facilitate harmonization and assignment of data governance responsibilities. We place a business information model in the center — used end-to-end from analysis, design, development, testing to data quality checks by data stewards. In addition, we show that the approach is suitable beyond traditional data warehouse environments, applying it also to big data landscapes and data science initiatives — where business requirements analysis is often neglected. As proper tool support has turned out to be inevitable in many real-world settings, we also discuss software requirements and their implementation in the Accurity Glossary tool. The approach is evaluated based on a large banking data warehouse project the authors are currently involved in.