MIGRATING TO AN OBJECT-ORIENTED DATABASE USING SEMANTIC CLUSTERING AND TRANSFORMATION RULES

This paper presents a methodology for handling an important step of da tabase migration. The methodology is based on a set of techniques: (i) semantic clustering, (ii) metamodeling, and (iii) knowledge-based sch ema transformation. Semantic clustering (i.e., grouping based on seman tic cohesion) is mainly used to facilitate the process of translating an extended entity relationship schema into a schema of complex object s. Meta modeling is used to define the data models involved in the pro cess of schema transformation. Finally, transformation rules are defin ed and used for mapping a schema from a source model into a schema exp ressed in a target model. In this paper, we limit ourselves to the map ping of an extended entity relationship diagram into an object-oriente d database schema using the object model supported by the ODMG standar d. (C) 1998 Elsevier Science B.V.