AAG Annual Meeting 2001 - Visualization Sessions

Session Title: Visualization II - Spatialization
Session Number: 4.2.15

Time: Thursday, Feb 29, 10:00 AM
Organizers: Sara Fabrikant and André Skupin

Specialty Group Sponsors: Cartography, GIS, Environmental Perception & Behavioral Geography

Christopher A. Badurek, Department of Geography, University at Buffalo, Buffalo, NY 14261. E-mail: badurek@geog.buffalo.edu.
Spatialization for Accessing Geospatial Images.

Currently, there is substantial interest in research initiatives that focus upon visualization techniques for accessing and finding meaningful geospatial information in very large databases. Visualization techniques that enable users to extract meaningful information from large data warehouses are increasingly important to the rapidly developing field of geospatial data mining and KDD (Knowledge Discovery in Databases). In particular, methods for representing and accessing the rapidly increasing amount of remote sensing and geospatial data need to be integrated. This paper presents an experimental design utilizing the visualization technique of spatialization for accessing geospatially-oriented images. The retrieval of images has long been a difficult task due to their subjective nature and a lack of research on information seeking behavior focusing on images. This paper aims to provide a framework for fusing the concepts of cognitive image schemata and geospatial metaphors into a digital representation of imagery and geospatial data for the purposes of geospatial information retrieval and data exploration. The object and data conceptualizations of images will be presented in terms of their relevance to geospatial cognition. This paper aims to provide a basis upon which an extensible geospatial imagery and information retrieval system can be constructed for large databases as well as examine cognitive aspects of spatialization of geospatial images.
Keywords: Spatialization, Visualization, Images

Sara I. Fabrikant, Department of Geography, University of California Santa Barbara, Santa Barbara, CA 93106, Email: sara@geog.ucsb.edu.
Regionalization and Scale-Dependence in Semantic Information Spaces.

This paper explores the geographic concepts of region and scale in semantic spatialization. Region and scale are not only elementary concepts in geographical analysis, but they are also associated with cognitive and experiential properties of the real world. Furthermore, scale becomes important when graphically representing a spatialization, as it affects the amount of detail that can be shown in the representation.
Computational methods for creating semantic information spaces involve a mathematical transformation that creates a logically defined coordinate system to re-arrange a set of data items based on their content and functional relationships. Most transformations are a variant of ordination, such as multidimensional scaling. MDS, however, collapses all semantic (intrinsic) properties of data object into the extrinsic property of spatial proximity, usually based on Euclidean distance. Topology (functional distance) and multiple levels of detail (hierarchy, functional regions) that are important aspects of the experiential world are not preserved with MDS. Visual exploration of the information space at different levels of granularity is therefore limited. Moreover, the reduction of detail that occurs as a result of the transformation procedure with MDS cannot be systematically evaluated.
The Intramax analysis on the other hand, allows to reveal scale-dependent functional regions in the information space, by hierarchically aggregating spatial units based on a functional distance measure. Intramax is a hierarchical clustering procedure that provides documentation and evaluation measures on how the regions are formed. This topological information helps to identify the appropriate levels of granularity at which the information space can be explored. The amplification of MDS with the Intramax procedure is a useful strategy for creating scale-dependent information spaces, that facilitate the exploration of abstract, complex data domains archived in rapidly growing digital repositories.
Keywords: spatialization, semantic information spaces, visualization

Michael Hermann and Heinrich Leuthold, Department of Geography, University of Zurich, Switzerland.
Spatialization and Spatial Metaphors in Social Science: The Combination of Qualitative Textual Analysis and GIS for Construction, Analysis and Visualization of the Ideological Space of Switzerland.

Our aim is to analyze and visualize spaces that represent the social differentiation of a society ap-plying the concepts of landscape and map. A major focus of the spatialization of social phenomena is the reconstruction and identification of the spatial dimensions, which are significant because they mark cleav-ages like social, cultural, political or mental conflicts. This makes it necessary to combine techniques of spatialization and visualization with qualitative research and interpretation embedded in sociological the-ory. Using the results of Federal Swiss Popular Referendums within the last 20 years, the authors recon-structed a space of mentalities and ideologies in Switzerland, the so called „Raum der Weltanschauungen". On this topic we show spatialization according to theoretical concepts from sociology and political science. For the naming and the identification of the extracted dimensions qualitative text analysis was combined with the quantitative analysis. The social relevance and meaning of the dimensions could be proven by the integration of socioeconomic and cultural indicators. Modeling and visualizing the ideological landscape with GIS allows very fine analysis and leads to the recognition of hidden social structures and regional patterns of ideological conflicts within the Swiss society. A comparison of social and linguistic diverse regions in the geographical space with their corresponding spread in the metaphorical space shows the in-fluences of these factors on regional political culture.
Keywords: spatialization, social spaces, cartography

André Skupin, Department of Geography, University of New Orleans, New Orleans, LA 70148. E-mail: askupin@uno.edu.
Cartographic Design for Map-Like Visualization of Information Spaces.

Information visualization is a rapidly growing field, which attempts to give people access to vast amounts of data by utilizing their spatio-cognitive abilities in a more direct way than previous information retrieval systems. These efforts are driven by such problems as information overload and disorientation. Many of the proposed solutions are based on the provision of map-like displays of non-geographic information. Researchers involved in this work commonly face such issues as graphic complexity or feature labeling, and have recently defined scale to be 'the dominant problem in information visualization'. Cartographic terminology and concepts abound in academic and commercial writings on information visualization, yet they are typically void of references to actual cartographic research. On the other hand, the involvement of geographers in the broader visualization community is typically restricted to geographic visualization efforts. This paper gives an overview of various contemporary spatialization solutions, contrasted with traditional cartographic examples as well as examples of the authors' own spatialization experiments. The computational techniques, data structures, and design decisions underlying these spatialization approaches are presented. The choice of particular design solutions is interpreted from a cartographic perspective. Particular attention is paid to future prospects of cartographic involvement in non-geographic information visualization.
Keywords: spatialization, cartography, map design

January 22, 2001