Revision with unchanged content. In this thesis we searched for ways to enhance human accessibility to the current Semantic Web technology by enabling the visualization of knowledge. We built a graph-based user interface in which users can explore and adapt information that is arranged with respect to the relations within this information as well as its meaning. The data necessary to arrange information that way is extracted out of RDF descriptions. The work in this thesis was done in the context of the SmartWeb project at the German Research Center for Artificial Intelligence (DFKI). The goal of the SmartWeb project is to lay the foundations for multimodal user interfaces to distributed and composable Semantic Web services on mobile devices. Due to the use of a mobile handheld device, our graph-based user interface must be able to present content on a small screen size. For the space-efficient visualization we use a graphical fisheye view for graphs in combination with advanced interaction forms to offer details on demand. The use of a graphical fisheye view is a valuable tool for seeing both "local detail" and "global context" simultaneously.