Many organizations start creating structured business glossaries. In particular, the business information modeling approach is followed to achieve proper data definitions and their harmonization across business areas. Creating a business information model from scratch is a significant effort. Existing artifacts in a structured (database schema, interface definitions) or an unstructured form (documentation, process descriptions) can be used as a starting point. In this thesis, we focus on processing the structured information.
The aim of the thesis is: (i) to review the state of the art in reverse engineering of data model and ontology learning, (ii) to design and implement a tool to extract a data model from a relational database and (iii) to create a business information model or update an existing one. The implementation will be used in the Accurity Glossary tool.