Building a Linguistic Model for Conceptual Data Model Quality

摘要

In this work, we present a new technique to obtain a linguistic model starting from a set of data. This model discovers and shows the. relationships that exist between some control variables and another objective variable in which, we are interested. Using the theory of Fuzzy Sets and by means of successive divisions of the definition space of the control variables, we generalize the induction algorithm CART to obtain a, set of linguistic rules thai reflects the set of well-known data. We show how we have, applied this tc.elmi.quc to find a linguis-tic model that reflects the existent relationships and dependencies between a set of metrics for entity-relationship diagrams (ERD) and the ERD maintainability, which is one of the most critical quality characteristic in the software development area. This model also allow us to predict ERD maintainability