The concept of spatial decision support has featured prominently in the GIScience literature of the last two decades for a simple reason: much of geospatial data processing is done to obtain information for decision making. Since almost any spatial information system can be claimed to offer some form of decision support, an effort was made in the late 1980s and early 1990s to define a minimum set of functionalities required of spatial decision support systems (SDSS). SDSS were originally proposed to aid individual (Densham and Armstrong 1987) and group (Armstrong 1993) decision makers in solving spatial, semi-structured problems, in which location and spatial relationships of distance, direction, connectivity, and adjacency are an integral part of problem solution, and decision objective(s), decision alternatives and their outcomes, and evaluation criteria are not fully known. Accordingly, the original purpose of SDSS was to assist individuals and groups in articulating decision objectives and evaluation criteria, finding feasible decision alternatives, and identifying superior decision options. To achieve this purpose SDSS needed to integrate: (1) mathematical and logical formalisms (models) to process spatial data and information, (2) technical data (scientific measurements), and (3) perceptual data (estimates, probabilities, and human judgments). These requirements mirrored the requirements for decision support systems (DSS) formulated by Keen (1977) and Sprague and Watson (1986) in the field of management science. Similar to DSS, a blueprint for SDSS was based on the idea of providing an easy access to spatial data and decision models through the integration of spatial databases, analytical models, and visualization tools (Densham 1991).
The above notion of what constitutes decision support has not been uniformly shared. For some, any data processing activity resulting in information of direct or indirect value to problem solving and decision making is synonymous with decision support. This broad understanding of the meaning of decision support reflects the trend to treat almost any computerized data processing system as a "decision support system" (Keenan 1997). In the same vein, Geographic Information Systems (GIS) have been portrayed as SDSS on the grounds that a GIS provides spatial data access, analytical processing, and display capabilities (Cowen 1988). Arguably, the computational basis for spatial decision support was then and still is today some form of GIS technology. Just like early SDSS, today's SDSS have at their core the basic decision aid(s) of GIS including data management to help extended human memory, graphic display to enhance visualization, and spatial analysis functions to extend human computing performance. Over the course of the 1990s and in recent years these common decision support functions have been expanded to include optimization (Aerts et al. 2003, Church et al. 2004), simulation (Wu 1998), expert systems (Leung 1997), multiple criteria evaluation methods (Malczewski 1999, Thill 1999, Feick and Hall 2004), on-line analysis of geographical data (Bedard et al. 2001), and visual-analytical data exploration (Andrienko et al. 2003) to generate, evaluate, and compute trade-offs among decision alternatives.
While much of the work on SDSS has been aimed at the design and development of software tools, another direction, started in the mid-1990s, has broadened the focus of research on SDSS by evaluating the usability and socio-behavioral consequences of decision support tools in the context of simulated (laboratory) and realistic (field experiment) decision situations (Reitsma 1996, Stasik 1999, Jankowski and Nyerges 2001a,Haklay and Tobón 2003). Some of this research was inspired by work in the field of human-computer-interaction (HCI) centered on collecting empirical data about the use of software tools by humans in order to explain why some software systems and techniques are more usable than others (Nyerges 1993).
Graphical Ontology Browser
- Click on a node to jump to the content of that node
- Pan to see the rest of the graph
- Scroll the mousewheel up and down to zoom in and out
- Rearrange the nodes in the graph by dragging a node to a different position