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

An adaptive interaction architecture for collaborative GIS.

Article Excerpt
Introduction

The research presented in this paper introduces context-aware mobile GIS that integrates adaptive interaction principles. We define an adaptive GIS as a generic GIS that can be automatically updated according to several contexts defined by (1) the properties and location of the geographical data manipulated; (2) the underlying categories that reflect different group profiles; and (3) the characteristics of the computing systems and supporting web and wireless techniques. This classification has been inspired by a related work by Calvary et al. (2003). The contexts correspond to the components that influence, to a certain degree, the diffusion of geographical data in wireless environments. The dimensions identified are of different nature as they involve geographical data, computing processes and interfaces, and user categories. These dimensions are not new when studied individually, but they are rarely considered as a whole.

For instance, previous works in the field of adaptive GIS introduce a technology-driven approach for a hardware-based interaction medium (Hampe and Paelke 2005; Burigat and Chittaro 2005). The adaptation of an open GIS layer descriptor to specific user needs and contexts has also been studied by Zipf (2005). A context-sensitive model for mobile cartography that emphasizes different levels of data adaptation and presentation has been proposed by Reichenbacher (2003). In the collaborative domain, a GIS is commonly used as a dialog and interaction medium between different users (Rauschert et al. 2002; MacEachen and Cai 2006). Implicit collaboration between several users may be used to recommend an interface content (Brown et al. 2002). These works implicitly show the diversity of the notion of context and the need for an integrated approach of the problem.

Similarly the development of personalized software has been the object of considerable attention by the information society (Riecken 2000). Web information engineering offers promising solutions for personal content profiling. Many different algorithms have been tested and are currently used by e-commerce web sites (Schafer et al. 1999). A key issue in modeling user preferences is to approximate user intentions with a few information inputs. The techniques used for extracting user preferences and categories vary from explicit user feedbacks (Shearin and Lieberman 2001), where a software agent learns user interests by interacting with her/him, to implicit tracking of user actions where preferences are deducted from her/his actions (Yang and Claramunt 2005). A promising direction currently explored takes a functional point of view where users are categorized according to their behaviors (Oard and Kim 2001). Most of these approaches are based on the fact that users are likely to share interests when they belong to a same category. A general drawback of existing methods is that no assumption can be made regarding a new user entering a profiling system. A user passes through a "cold start" period of "undetermination" while her/his profile is partially defined (Schein et al. 2002). Profiles may not immediately reflect shifts in user behavior, which is manifested by an inappropriate inertia of a user profile. This "profile inertia" problem was emphasized in the domain of web service personalization (Lam et al. 1996).

[FIGURE 1 OMITTED]

Maritime navigation encompasses several properties that form a relevant experimental context for the development of a collaborative and adaptive GIS: user mobility, spatial distribution of services, and variability of user needs. In a previous work, we introduced an architecture and real-time services for the diffusion of maritime geographic information, at different levels, from the global monitoring of the maritime traffic of a given area (Desvignes et al. 2002) to individual services on request (Dubs and Kaufmann 2006).

The objective of the research presented in this paper is to consider the environmental knowledge of a distributed GIS as the input of a collaborative adaptation process that avoids the "cold start" and the "profile inertia" problems. We consider the way a given user acts in the environment as a contextual input of an adaptive GIS. The characterization of user behaviors and aggregation of users into groups of similar properties aim to expedite the learning process of novel users. The proposed interfaces and contents are adapted to a given situation, and shifts in behaviors and data demands are reflected by a regular derivation of profiles within each group.

The remainder of this paper introduces a modeling approach applied to distributed GIS and experiments with it in the context of maritime navigation, an emerging field of GIS that combines mobility and distributed services. Section two presents a framework for the integration of contextual dimensions into a collaborative GIS. This is followed by an introduction and modeling of the notion of geographical context. The next section describes the principles that underline the concepts of user groups, and their derivation mechanisms. The last but one section introduces collaborative interactions at the group level and how these serve as a support for a generation of adaptation profiles. Finally, in the last section, we draw some perspectives.

Adaptive GIS Framework

The collaborative component of the adaptive GIS framework is derived from user behaviors and patterns that allow for the generation of different users groups. The computing and interface environment considered are the one of a mobile and distributed GIS, considered as an adaptive GIS.

We characterize such an adaptive GIS by three contextual dimensions (Figure 1). An adaptive GIS should be able to automatically derive its content and interface from a changing environment. This implies ability to present relevant information to the user and to improve the usage and usability of the information provided. An adaptive GIS is built upon several services that integrate geographical data and services. Our framework considers that an adaptive GIS is built on one-to-many location-based services, preferably bounded by a common user interface and delivering geographical data, visualization, and querying facilities. The context of a mobile GIS--with respect to a given user acting in the environment--varies according to the status of the different contextual dimensions considered. Depending on the current geographical context (i.e., what and where), the facilities offered by a service may be available or not. Available facilities can be refined by taking into account user preferences (i.e., who) and application capabilities (i.e., how). These contexts form the core of an adaptive...

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