UML to OWL Mapping Overview An analysis of the translation process and supporting tools David de Almeida Ferreira 1 , Alberto Manuel Rodrigues da Silva 2 . 1) INESC-ID/Instituto Superior Técnico, Lisbon, Portugal david.ferreira@inesc-id.pt 2) INESC-ID/Instituto Superior Técnico, Lisbon, Portugal alberto.silva@acm.org Abstract This paper presents a brief overview of the translation techniques between the OMG’s UML modeling language and ontological engineering languages such as the W3C’s standard for Semantic Web, the Web Ontology Language (OWL). The main objective of this paper is to analyze the feasibility and easiness of modeling an ontology by using common and mature UML CASE tools, thus reducing the gap between software engineering practitioners and AI techniques for ontology creation. The analysis and practical application of a proposed MOF-based Ontology Definition Metamodel (ODM) and the respective Ontology UML Profile (OUP) allows the straightforward comparison between MOF and OWL concepts. Keywords: Modeling, Translation, Semantic Web, Ontology, UML, MOF, ODM, OUP, OWL. 1 Introduction Nowadays we are observing a significant effort of introducing meaning (semantics) to data stored in the Internet. The W3C started to develop an ambitious framework called “Semantic Web” to define common formats for knowledge data interchange for overcoming the original Web limitation of only interchanging data through documents [Miller 2004], which constitute a coarse-grain level of knowledge bundles. To achieve this goal, the Semantic Web effort also engages the definition of a XML-based language for recording how data relates to real world objects allowing a human or a machine to store, access, and interchange that knowledge with a wide range of databases, which are connected through a lattice of concepts. The main objective of the Semantic Web framework is to enable data to be shared and reused across application, enterprise, and community boundaries in a straightforward manner. Domain ontologies are the formal organization of domain-specific knowledge, thus constituting the most important part of Semantic Web applications. Besides the advantages in terms of knowledge interchange through heterogeneous environments, the developed ontology-based framework reduces the gap between the wider software engineering population and the disregarded, or even unknown, potential of Artificial Intelligence (AI) techniques, like reasoning. The main purpose of this paper is to gather information about the current state of the art in this investigation area, and present some initiatives that allow one to create an ontology by using Unified Modeling Language (UML) CASE tools, and further refinement with ontology-specific tools. One of these initiatives, which provided a strong and consolidate theoretical background on the field, proposes an Ontology Definition Metamodel (ODM) [Djuric 2004]. This MOF-based