| The multi-lingual database system serves as a test-bed for examining data-model transformations, data-language translations and data-model semantics. The different data-model transformations are used to study the data-model semantics. We are comparing and contrasting different data models from a data-semantic viewpoint, by examining their data-model transformations to a common data model. For the data-language translations, we develop a method for characterizing the translations. We then use this method to compare different data languages, as they have been translated into a common data language. Thus, this dissertation includes the processes of data-model transformations, data-language translations and their relationships to data-model semantics.;In order to compare different data languages, we have developed a compact and uniform method for specifying the data-language-translation process. The specification method is used for comparing the different data languages. The end result of this analysis is the determination of the different types of data languages through the data-language-translation process.;In this dissertation we have studied how four different data methods (i.e., the relational, the hierarchical, the network and the functional models) can be transformed to a common data model (i.e., the attribute-based model). Using the quantitative measure and qualitative analysis for the data-model-transformation processes we have determined a ranking of the four different data models. We have also studied how the corresponding data languages (i.e., SQL, DL/I, CODASYL-DML and Daplex) can be translated to a common data language (i.e., ABDL). Using the specification method for the data-language-translation processes we have determined the types and characteristics of the different data languages. Collectively, through this dissertation, we have learned more about data models, data languages and data-model semantics.;In order to compare and contrast different data models, we have developed a quantitative measure and a qualitative analysis for examining the transformations. The quantitative measure gauges the complexity of the data-model-transformation processes. The qualitative measure compares the data models: first, by characterizing each data-model transformation, and second, by comparing these characterizations. Both the quantitative measure and the qualitative analysis give us a means for investigating data-model semantics and establish a ranking of the different data models. |
