Context comparison for object fusion

Abstract

We propose a solution to help the integration of heterogeneous sources based on the fusion of objects according to their context. New requirements for information exchange have emerged with all the developments around Internet. Information consumers want to access and combine data from remote and heterogeneous sources in a transparent and dynamic way. To achieve this level of interoperation is yet a real challenge. We present a model to define local data as informative objects with a contextual representation associated to them. A semantic context comparison mechanism, based on a semantic distance, reconciles context of applications and constructs virtual objects in which rules make the fusion of informative objects. Virtual objects play the role of interface for transparent user access. © Springer-Verlag Berlin Heidelberg 2003.