Abstract
How to enable enterprises to interchange autonomously designed information addressing the same domain, is an active research area. Present technology allows enterprises to exchange data, but when it comes to interpreting the content or understanding the data, there is a problem. This problem has to do with semantic interoperability; enabling computers to understand the meaning of information being exchanged.
Two recent initiatives make promises regarding interoperability issues. Ontologies have been adopted by the field of Artificial Intelligence (AI), and play an important role when envisioning a computer-understandable and executable Web referred to as the Semantic Web. An ontology and related technology is to directly address semantic issues by providing an explicitly defined meaning of available services and the information to be exchanged. The Model Driven Architecture (MDA) on the other hand, addresses interoperability by addressing the conflicts introduced due to the different technological platforms chosen when implementing systems. The MDA framework enables developers to carry out their work at a higher abstraction level as well as applying model transformations to automatically generate technology-dependent parts of systems.
Semantic mapping denotes the task of capturing relationships between semantically similar terms in different data sources as well as supplying the techniques needed to convert between them. This thesis investigates how ontology is used in relation to semantic mapping with emphasis on what distinguishes the ontology-based approach from the MDA approach. An effort is made to explain what ontology denotes within the field of AI as well as to compare the ontology technology used to construct and apply ontologies, to the model transformation technology available within the MDA framework. This thesis is also meant to give the reader a good understanding of the challenges involved when trying to achieve semantic interoperability between enterprises.
The main contribution of this thesis is the account of similarities and dissimilarities between the ontology-based approach and the MDA approach to semantic mapping. Important questions are whether those two are alternative approaches, or whether they may be combined in a value-adding way.