A Semantic Oriented Method for Conceptual Data Modeling in OntoUML Based on Linguistic Concepts

摘要

Conceptual data models, as means of communication, must have semantic quality. Such quality relies on the model's completeness and validity in relation to the concepts it is supposed to represent. Since the modeler acquires such concepts mostly from texts created in a natural language, a semantic-oriented linguistic approach should be adopted for building unambiguous conceptualizations. Also, the chosen modeling language must offer enough constructs for the creation of a faithful representation, like OntoUML. Such languages, however, may require a learning period that modelers hardly can afford. This paper proposes a modeling method that consists of systematic steps to promote the understanding of the concepts inherent to the domain to be modeled. The method application is illustrated in an example. Additional evaluations of the proposed method included a case study, which results indicated that it makes modeling less complex by allowing for modeling choices to be dealt with within the realm of the natural language.