The exponential growth of the World Wide Web in the last decade brought an explosion in the information space, which has important consequences also in the area of scientific research. Finding relevant work in a, particular field and exploring the links between publications is currently a cumbersome task. Similarly, on the desktop, managing the publications acquired over time can represent a, real challenge. Extracting semantic metadata, exploring the linked data cloud and using the semantic desktop for managing personal information represent, in part, solutions for different aspects of the above mentioned issues. In this paper, we propose an innovative approach for bridging these three, directions with the overall goal of alleviating the information overload problem burdening early stage researchers. Our application combines harmoniously document engineering-oriented automatic metadata extraction with information expansion and visualization based on linked data, while the resulting documents can be seamlessly integrated into the semantic desktop.
