Ontology Alignment through Argumentation

Currently, the majority of matchers are able to establish simple correspondences between entities, but are not able to provide complex alignments. Furthermore, the resulting alignments do not contain additional information on how they were extracted and formed. Not only it becomes hard to debug the alignment results, but it is also difficult to justify correspondences. We propose a method to generate complex ontology alignments that captures the semantics of matching algorithms and human-oriented ontology alignment definition processes. Through these semantics, arguments that provide an abstraction over the specificities of the alignment process are generated and used by agents to share, negotiate and combine correspondences. After the negotiation process, the resulting arguments and their relations can be visualized by humans in order to debug and understand the given correspondences. The existence of heterogeneous data models in computer systems leads to an integration problem when two or more of these systems need to interact and exchange information. This can be due to several reasons, including differences in model representation languages, structure, constraints and semantics, where the origin is often because of a lack of consensus (Sheth and Larson 1990) between those who built the models. Model matching, which consists in finding correspondences between the entities in both representations (or models), is considered to be the first step in solutions for information integration (Euzenat and Shvaiko 2007). With the increasing popularity of the Semantic Web, more and more data models are being published daily in the form of ontologies. This increase in the amount of models and their heterogeneity is becoming a global scale integration problem. Even so, the demand for complex ontologies in the Semantic Web is small. Actually, empirically, there seems to be a struggle to create very simple and easily shareable and reusable ontologies (as they can more easily become a consensus). However, in the case of business enterprises (Silva, Silva, and Rocha 2011) and in specific research domains such as genetics (Goble and Wroe 2004), complex and heterogeneous ontologies exist. When such ontologies need to be aligned, matches can involve different types of Copyright c 2012, Association for the Advancement of Artificial