Automatic Discovery of Association Paths in Relational Databases Using Software Visualization

Abstract

We introduce a visual framework for facilitating tasks associated with database maintenance and re-use. The prototype system embodying the framework is presented. The system utilizes various techniques and features of software visualization. The system supports visual displays of the database structure along with various implicit relationships found in it such as associations and path views. Information visualized is automatically extracted from the database schema. To assess the usefulness of the proposed framework in helping the programmers to quickly recognize path views among relations, an empirical evaluation was conducted. Results collected from the evaluation seem to support our hypothesis that the time required to manually recognize path views from the database schema is considerable and tends to increase as the depth between the relations increases. The evaluation also showed that by using our visual framework such time is negligible and tends to be static.