F-ALCI : A Fully Contextualized, Federated Logic for the Semantic Web George Voutsadakis 1,2 , Jie Bao 3 , Giora Slutzki 1 , and Vasant Honavar 1 1 Department of Computer Science, Iowa State University, Ames, IA 50011 2 School of Mathematics and Computer Science, Lake Superior State University, Sault Ste. Marie, MI 49783 3 Department of Computer Science, Rensselaer Polytechnic Institute, Troy, NY 12180 Abstract. In this paper, a new version F-ALCI of the package-based counterpart ALCIP - of the description logic ALCI is introduced. It allows contextualization of all the logical connectives rather than just logical negation. Moreover, a new semantics is introduced that is based on image domain relations, but is not ridden with overly restrictive con- ditions from the outset. One may impose additional conditions if stronger properties are required in a specific framework. These features allow more flexibility and generality in the modeling process. To show that this con- textualized federated description logic is decidable, a sound and complete reduction to the description logic ALCI is provided. 1 Introduction Recent efforts aimed at enriching the world-wide web with machine interpretable content and interoperable resources and services, are transforming the web into the semantic web [7]. The semantic web, much like the world-wide web, relies on the network effect, that is on leveraging the work of independent actors who contribute resources that are interlinked to form a web of resources. In short, web pages : web :: ontologies : semantic web. The ontologies that provide a basis for establishing the intended semantics of resources (databases, knowledge bases, services) that constitute the semantic web are typically developed independently to serve the needs of specific communities. They typically cover different, par- tially overlapping, domains of discourse (e.g., biology, medicine, pharmacology). Inevitably, the axioms that make up the ontologies are applicable within the contexts that are implicitly assumed by their authors. However, many appli- cation scenarios require selective use of knowledge from multiple independently developed ontology modules. For example, a group that is focused on translating discoveries that link genetic and environmental factors to specific diseases into effective therapies might need to selectively reuse the contents of an ontology created for use in one context (e.g., genetic studies) in a different, but related context (e.g., drug design). Reaping the benefits of the network effect in such a setting requires theoretically well-founded yet practically useful approaches