| Abstract |
In this talk I will take an abstract, model-theoretic view on the problem of representing and reasoning with contexts in some common terminological languages and knowledge representations, such as used on the Semantic Web. I will sketch a conservative extension of their current, Kripkean semantics, which arguably offers a very powerful and intuitive framework for representing and integrating contextualized information. The extension is well-established in the world of mathematical logic (many-dimensional modal logics) as well as in AI (J. McCarthy's theory of contexts). Finally, I want to mention a few issues concerning the complexity and expressiveness of the languages capable of exploiting such a semantics. I will be very interested in obtaining a feedback on how the presented perspective seems to fit into the landscape of the relevant problems and challanges in the real-life applications on the Semantic Web. |