Dr. Antonio Badia

Dr. Antonio Badia is an Assistant professor in the Computer Engineering and Computer Science department at the University of Louisville, where he directs the Database Lab and the Data Mining Group. He holds an undergraduate degree in Philosophy and got his Phd in Computer Science in 1997 from Indiana University. His research interest includes query optimization, conceptual modeling, active databases and database tools for knowledge management. His current research is supported by the NSF and the U.S. Navy.

ABSTRACT:

Requirement Specification is a first, vital step in the development of large software systems. In the case of Information Systems, the most important activity during this step is Conceptual Modeling. During Conceptual Modeling, a part of reality under consideration (the domain of discourse) is analyzed and systematically captured in a {\em conceptual model}, a semi-formal specification of how to structure and express information. Two of the most successful and widely used conceptual models for Information System design are the Entity-Relationship (E-R) model and the Unified Modeling Language (UML) class diagram. They are both relatively simple, intuitive, yet able to express the most common situations. There are also well known methods to design databases starting from either one. It would seem, then, that specifying databases is a well understood, solved problem. However, when we scratch the surface, serious issues arise. Both models have severe expressive limitations. These in turn result in limitations on the design phase. We show that these limitations seriously restrict the functionality of the resulting database. We then review the role of conceptual models and suggest avenues of research to fix this situation.