Graph-based privacy preference expression for the semantic web

Abstract

The Web is changing. Originally a medium for human-readable documents, the next generation Semantic Web is opening up toolkit that provides vast opportunities for sharing information that is encoded in a machine-readable form. The opportunities for sharing data also are opportunities for encroaching on people's privacy. While some technologies already exist for expressing privacy preferences, they do not integrate closely with the Semantic Web. In addition, previous approaches to privacy expression do not easily expand to data shared over multiple hops, with different privacy preferences at each hop. The Private Information Management Agent is an attempt to mitigate these concerns.