CODASYS

Abstract

The paper describes the main features of a prototype tool CODASYS (COnceptual modeling tool for DAtabase SYStems), which purports to help novice designers engaged in conceptual data modeling. It is well known that conceptual data modeling is an error-prone process for novice database designers. The tool assists a designer in developing an entity-relationship diagram that can be translated to a normalized relational representation, free of derived dependencies. We first discuss a set of requirements for the tool that is based on achieving normal forms, on preventing data modeling errors, and on a theoretical foundation. We then discuss how the tool achieves the four normal forms and prevents common database errors such as incorrect degree, incorrect connectivity, and derived relationships. Further, we provide a cognitive framework for understanding novice error behavior and elaborate on how CODASYS addresses cognitive strain in database design. Next, we present the features of the tool and, with the help of flowcharts, discuss how these features have been implemented using a rule-based approach. It may be noted that empirical studies have been conducted to test the efficacy of the tool and have indicated significant improvement in novice designer performance.