Home | Business News | Browse by Publication | C | Cartography and Geographic Information Science

Reconciliation of inconsistent perspectives in collaborative GIS.

Article Excerpt
Introduction

In the early days, geographic information systems (GIS) were used by individual as a tool to collect, store, manipulate, and visualize spatial data and to provide spatial analysis functions (Demers 1997). Later, the computer-supported cooperative work (CSCW) concept was integrated with GIS to support geospatial collaboration among multiple people (Li and Coleman 2002). The use of GIS to provide a cartographic display media in group work with geographic information (Armstrong et al. 1992; Cai et al. 2005), and the integration of GIS with CSCW to support collaboration and coordination, have developed into a broad body of research generally referred to as collaborative GIS (Chen et al. 2004; MacEachren and Brewer 2004). Collaborative GIS has widely been applied but not limited to support group spatial decision-making (Armstrong 1993; Jankowski et al. 1997; Jones et al. 1997; Fuhrmann et al. 2003; Aggett and McColl 2006). In Changzhou Municipality (Jiangsu Province in China) for instance, a GIS-based CSCW system has been implemented to provide support to collaborative urban planning and land management (Jiang and Chen 2002). Such implementations of collaborative GIS involve many participants with different cultural and educational backgrounds and personal experiences and preferences (Malczewski 2006). Their perspectives of decision-making issues may be inconsistent and need to be negotiated. This paper focuses on modeling a reconciliation process aimed at achieving consensus before selection.

Collaborative GIS has spatial and temporal characteristics, i.e., it can be of a same or different place and of a same or different time (MacEachren 2001). The Changzhou application is an example of a same-place geospatial collaboration application, with both same and different time groups of people making urban planning decisions in the 1990s. The urban planning process comprised several steps, usually beginning with the processing and a review of building/land-use applications submitted by public agencies or private citizens. This work is implemented by urban and land management planners who are usually staff members of the same agency or bureau. They belong to different departments, but they can meet easily. If there is a disagreement with the perspectives proposed by others, these perspectives are returned for modification, or conferences may be organized where inconsistent perspectives are discussed face to face. The latter is a direct communication process which is not modeled in same place applications. In same place applications, people pay much attention to the coordination between the system and users. The methods employed in system-user negotiations to achieve consensus understanding based on a shared pool of knowledge have been the subject of in-depth research that yielded the collaborative discourse theory (Grosz and Sidner 1986), the recipe graph (or Rgraph) which describes the intentional structure of a collaborative discourse (Lochbaum 1998), and the collaborative dialogue approach which was extended to represent vague spatial concepts (Cai et al. 9003).

In some situations, people are located in different geographical places and they cannot meet easily. For example, when in 9003 China suffered a high incidence of Severe Acute Respiratory Syndrome (SARS), people's movements were restricted to avoid infection with, and transmission of, the virus, highlighting the need for a system that would support collaboration among the home-bound participants of the decision-making team. The distributed collaboration GIS emerging in recent years satisfies that need (Dragicevic and Balram 2004; Bapna and Gangopadhyay 2005; Balram and Dragicevic 2006a) by enabling the exchange of personal views on given topics, irrespective of location. The two most popular of such systems are Toulmin's argumentation model and the Issue-based Information System (IBIS). They describe the central elements and their relationships for structuring the argumentation process and employ a multi-criteria decision model and a group voting tool to select appropriate decisions (Pereira and Duckstein 1993; Jankowski and Nyerges 2001; Chakhar and Martel 2003). None of the models mentioned above does, however, describe the actions participants take to express their suggestions and the negotiation process people go through to achieve consensus in distributed environments (Balram and Dragicevic (2006b)).

This paper presents a method for reconciling inconsistent perspectives among participants in distributed collaborative environments. An extended Speech Act Theory (SAT) was utilized to help define and describe collaborative speech acts. A coordination process is established based on these speech acts, and the key techniques of the coordination process are researched. The paper is organized into six sections. The next section discusses the representation of inconsistent perspectives, the characteristics of reconciliation in distributed environment, and the framework of a proposed SAT-based reconciliation method. The following section defines the basic speech acts involved in reconciling inconsistent perspectives; then the coordination process is constructed based on these speech acts. We then describe the key steps used to reconcile different perspectives using this process. The proposed reconciliation method is illustrated on a case study, highlighting the experimental method and its results. The last section presents a discussion and some conclusions.

[FIGURE 1 OMITTED]

Framework of SAT-based Reconciliation Method

In GIS-supported collaborative applications, participants propose their perspectives on spatial decision problems and work together to reconcile inconsistencies in perspectives due to the collaborators' different backgrounds. The perspectives are usually represented as spatial objects with related attributes, such as regions to identify the spatial scope of these objects and polylines to denote the spatial route they take. The inconsistency of two perspectives, for instance, is akin to the difference between two spatial objects, defined by the objects' different spatial locations, shapes, and attributes. Figure 1 shows two examples of inconsistent perspectives, one based on different attributes (Figure la) and the other on a different spatial shape in the polyline (Figure 1b).

In distributed environments, the expression of perspectives and the coordination of inconsistency are implemented by participants through such spatial operations as adding a land-use...

View this article FREE - Now for a Limited Time, try Goliath Business News
Free for 3 Days!