A negotiation protocol for data integration driven by ontology

Abstract

Data integration comprises of a combination of data available in different sources, allowing users with a unified view of these data. After three decades of research aiming to resolve data integration problems, we are still unable to classify it in a single problem. Additionally, researches indicate that the use of formal ontology is a suitable resource for dealing with data integration problems and can contribute with open problems, such as: system architecture, content and functionality, data type, interoperability and performance requirements. The current environment for such integration is the World Wide Web, in which information and service are enhanced by semantic definitions. The Semantic Web is the new frontier, allowing the Web platform to understand and provide users and computers with a much richer content, as knowledge. Ontology can also contribute with knowledge acquisition issues, particularly on the Web, since it is expert dependent, tacit knowledge to be considered and expert availability. Ontology must be available, in the form of a modular knowledge base, to guide its acquisition, allowing reuse and sharing. Regarding data complexity where knowledge is unstructured, representing different domains, we propose a framework that uses a negotiation protocol. The protocol identify one or more integration meaning points, and ensure certain level of semantic agreement, that is, two systems try to ensure that they agree on the knowledge needed. The result of the negotiation protocol becomes a set of knowledge which will be encapsulated and made available for use, as well as, part of the Web content.