Managing Master Data with XML Schema and UML

Abstract

To manage parameters, the majority of Information Systems is concerned by heterogeneity in both data and solutions. Consequently, the management of this data becomes complex, inefficient, insecure and expensive. The need to use a structured formalism to handle complex data appears. We suggest a data integration solution based on a XML architecture. This architecture embeds a Master Data Management in the Information System. The unification of Master Data is primarily done by the definition of models. These models are XML Schema documents describing complex data structures. We propose to enrich the structure and the semantics of these models by defining a metamodel. In the metamodel, we introduce semantic object relations for defining links between concepts. The resulting metamodel is used to define an UML profile and to optimize operations such as models validation, data factorization and trees representation. Moreover, UML profile is exploited to make easier the definition of models.