A transformational approach to GIS operations

GIS research has a continuing threaddevoted to classifying the operations performed by analytical software. Prior e€ orts to classify GIS operations have limitations and inconsistencies, often arising from an attempt to establish overly direct links between geographic procedures and arithmetical operations. The transformational view of cartography provides a more solid basis for classifying GIS operations. This paper presents a new scheme for geographical transformations based on measurement frameworks as the principal distinction. Transformations between measurement frameworks can be summarized in terms of a spatial neighbourhood and a rule to process attribute information. This scheme organizes most analytical GIS operations by their geometric and attribute assumptions. 1. Demystifying the conversion of data into information A geographical information system (GIS) converts raw data into more useful information (an anti-entropic feat) through a maze of analytical procedures that are poorly explained and often misunderstood. Most accounts place a strong emphasis on the analytical toolkit (embodied in software) as the agent of the conversion. The GIS is presented as a marvellous machine that can distil valuable information from huge stockpiles of ordinary data. Such a claim should not be propagated without documenting the mechanisms that enhance the content of a data source. While software can compute any number of new relationships, only a few of these hold potential value. Users of these software tools should be better informed of the applicability of various software tools to their particular situation so that the information serves the intended purpose. Some authors de® ne GIS around a speci® c range of tools to integrate diverse forms of geographic information (Cowen 1988, Maguire 1991). Yet, behind the grand claims, the procedures of integration vanish into a maze of special cases and an everexpanding list of software functions. Students and neophyte users are at a loss to navigate the maze we have built. Yet, despite the lack of articulate description, practical applications of GIS routinely manage to use the tools to create new information not inherent in the available sources. The goal of this paper is to present a taxonomy for GIS operations that is su ciently comprehensive and ̄ exible while reducing the maze to a clear structure. This paper will begin with a review of the Internationa l Journal of Geographica l Information Science ISSN 1365-8816 print/ISSN 1362-3087 online Ñ 1999 Taylor & Francis Ltd http://www.tandf.co.uk/JNLS/gis.htm http://www.taylorandfrancis.com/JNLS/gis.htm