Sharealike your data: Self-referential usage policies for the semantic web

Abstract. Numerous forms of policies, licensing terms, and related conditions are associated with Web data and services. A natural goal for facilitating the re-use and re-combination of such content is to model usage policies as part of the data so as to enable their exchange and automated processing. This paper thus proposes a concrete policy modelling language. A particular difficulty are self-referential policies such as Creative Commons ShareAlike, that mandate that derived content is published under some license with the same permissions and requirements. We present a general semantic framework for evaluating such re-cursive statements, show that it has desirable formal properties, and explain how it can be evaluated using existing tools. We then show that our approach is com-patible with both OWL DL and Datalog, and illustrate how one can concretely model self-referential policies in these languages to obtain desired conclusions.