Andrew Dillon wrote an interesting paper about shape and information space in Journal of the American Society for Information Science: Spatial-Semantics: How Users Derive Shape from Information Space
The ability to perceive structure in abstract information spaces is crucial to navigation and search performance. Dillon's article distinguishes the role of spatial and semantic cues and explains why this conceptualization may lead to new insights into existing and emerging data. Dillon also introduces the concept of shape as the structural component of the working model of an information space. This is most apparent in Geographical Information Systems (GIS) but is less obvious or conceptualized in abstract information environments. Dillon's article delineates the argument between top-down versus bottom-up approaches with a range of empirical evidence found in the literature.