Bry, François; Colonna, François-Marie; Lorenz, Bernhard
Multi-Paradigm Reasoning for Access to Heterogeneous GIS.
Accessing and querying geographical data in a uniform way has become easier in recent years. Emerging standards like WFS turn
the web into a geospatial web services enabled place. Mediation
architectures like VirGIS overcome syntactical and semantical heterogeneity
between several distributed sources. On mobile devices,
however, this kind of solution is not suitable, due to limitations,
mostly regarding bandwidth, computation power, and available storage
space. The aim of this paper is to present a solution for providing
powerful reasoning mechanisms accessible from mobile applications
and involving data from several heterogeneous sources.
By adapting contents to time and location, mobile web information
systems can not only increase the value and suitability of the
service itself, but can substantially reduce the amount of data delivered
to users. Because many problems pertain to infrastructures
and transportation in general and to way finding in particular, one
cornerstone of the architecture is higher level reasoning on graph
networks with the Multi-Paradigm Location Language MPLL. A
mediation architecture is used as a “graph provider” in order to
transfer the load of computation to the best suited component –
graph construction and transformation for example being heavy on
resources. Reasoning in general can be conducted either near the
“source” or near the end user, depending on the specific use case.
The concepts underlying the proposal described in this paper are
illustrated by a typical and concrete scenario for web applications.