Establishing an OGC Web Processing Service for generalization processes

Exchanging knowledge in terms of sharing and reusing generalization algorithms was identified as a key requirement by the generalization research community to ease further research (Edwardes et al., 2003). Therefore a common mechanism as Web Services, which provides an open platform for world-wide access to different functionalities, was suggested to be promising. Web Services are distributed over the Web and provide certain functionality, i.e. generalization operators of various developers. Additionally the application of Web Services for web-based generalization processes would benefit from the established web-based data dissemination approaches, which are mostly implemented by Spatial Data Infrastructures (SDI). Hence Web Services enable on-demand and on-the-fly generalization processes based on the most current data. By the way, different commercial software solutions (e.g. ArcGIS) incorporate only basic generalization functionality but utilize Web Services, thus enabling web-based generalization processes would also increase their functionality. However to apply Web Services as a means of knowledge exchange, these Web Services have to be accessible in an interoperable way based on common standards. Such a common standard (e.g. SOAP) has been applied by different research projects about web-based generalization processes (Neun and Burghardt, 2005; Neun et al., 2006; Regnauld, 2006). These projects gave already good insights into the capabilities of Web Services, but did not reflect the geospatial issues due to the missing geospatial concepts within SOAP (e.g. feature encoding). Thus some further research about evolving standards within the GI-domain has to be carried out. Currently a geospatial processing service interface is getting standardized by the Open Geospatial Consortium (OGC), labeled as OGC Web Processing Service (WPS) (OGC, 2005). Since the generalization process is spatially related, it is relevant to investigate possibilities and limitations of the WPS specification with respect to requirements of generalization applications and research. The following sections (2 & 3) will give a short insight into the WPS specification and its relevant features in terms of generalization processing. Section 4 presents the architecture of the developed WPS and shows how its implementation enables a research platform for generalization by addressing technical (extensible software architecture) and organizational (Open Source licensing) aspects. Sections 5 & 6 give an example of accessing web-based generalization functionality provided through the WPS and show the drawbacks of the current version of the specification. The last two sections describe the broader context of the related research and come up with a conclusion. All the presented results are part of an on-going research project at ITC’s Department for Geo-Information Processing (GIP) (Poppe et al., 2006).