Driven by the new information and communication technologies, geo-information services are more and more gaining attention in research, industry and administration. But how these services are designed is still an open debate. A common goal is to produce a cartographic product portrayed either on screen or printed on paper, with new properties, like interaction. We will present here some standards and frameworks which will lead us 'towards open geo-visualization services'.
With 'geo-visualization' we mean the graphic presentation (or portrayal) of geo-referenced, non-symbolized data (called 'geodata') which typically is being designed with an object oriented approach. A 'geo-visualization service' is able to deliver a map based on a request. If this service is targeted towards the web, we speak of a web-service, otherwise, it is a service dedicated for other specialized tasks. We mention 'open' because these services consist of components which have explicitly documented program interfaces based on freely available standard specifications. 'Open' also indicates new business opportunities especially also for cartographers.
We would like to start with ten theses about 'open geo-visualization services', five of them are problem statements the others indicate proposed solutions:
An important item to explain first is the 'model-based method'. This method means the application of data modeling and the usage of system-neutral and platform-indepenent interface services. Encoding rules automatically specify data structures and application programming interfaces can automatically be derived both from the underlying application models. Although this approach follows international standards (like ISO 19100 or the 'model-driven architecture' of OMG) and though data modeling is well established in database technology, researchers and users in cartography only recently began to exploit this method.
First, we mention implemented tools which are based on Unified Modeling Language (UML) and eXtensible Markup Language (XML). Then, INTERLIS 2 is being shortly introduced as a link between UML and XML.
After over ten years of experience in the implementation and use of this model-based method we have enough indication that it is feasible and that it is the crucial factor of successful geo-information services.
Since the publication of the Scalable Vector Graphics (SVG) specification there exists now a promising graphic vector output format which is a perfect match to our open standard INTERLIS.
Beginning of this year a feasability study was carried out in order to demonstrate the potentials of this model-based method. The goal was to generate a SVG output from INTERLIS base geodata (or almost any other format) according to a standardized graphic description. This graphic outcome should include standardized colored symbols as well as demonstrate simple interactivity by integrating hyperlinks.
Before we explain the results it is important to note, what is the role of SVG is in this standard environment. The basic visual presentation process is defined in figure 1. This abstract schema indicates roughly the data flow and the dependencies (dashed) starting with non-graphic geodata until it gets visualized on the graphic display as (carto-)graphic objects.
Figure 1: Graphic descriptions, on one hand built upon data and views, on the other upon graphic signatures.
We clearly understand graphic data, like SVG, as output or intermediate data. Any change in the data except either in the base data or in the standardized graphic description (a sort of legend and plotting configuration file) will lead to data sets which are hard to update.
The main work of the feasability study can be found on the right part of figure 1. A prototype "SVG writer module" was designed. A library from a commercially available geo-information web-service product called 'Geoshop' (see www.infogrips.ch) was chosen. A medium and a larger data set served as test geodata.
The prototype has been implemented in C in about a week by a couple of computer science students; it will be freely available at www.integis.ch. We report the main implementation problems with SVG we have encountered. A visual assessment of the resulting screen maps serves as a first evaluation of the prototype software and its outcome (see figure 2). Finally, we indicate some functional improvements of the writer module and the according graphic definitions in order to enhance both the graphic output quality as well as the functional features of the web map.
As a conclusion we can state, that - due to model-based INTERLIS technology - a very advanced geodata infrastructure is established in Switzerland ready for the broad implementation of open geo-visualization services!
Figure 2: Overview map of the first test geodata set.