GI-Days 2008 - 27/08/2008
GI-Days is an annual conference organised by the Institute for Geoinformatics at the University of Muenster that bringing together researchers from, mainly German, universities and research centres.<P>
This small-scale, two-day event for academics offers good insight into current geomatics research, and this year from 16th to 17th June provided fifty or so participants with a packed programme of presentations. Sponsoring companies presented their products and services at a small accompanying exhibition.
The web as platform for viewing, accessing and processing spatial data is a major theme in all research, and most projects are, remarkably, funded with money from the European Union purse. One example is SWING, which began in 2007 and deals with semantic web services interoperability for geospatial decision making and demonstrates the whole chain of semantic modelling and processing using geological data.
Many research papers are based on OGC specifications, especially the recently published WPS (web processing services). The ‘geospatial web’ is a recurrent research topic, defined by OGC as the complete integration and use of location at all levels of the web. Research on WPS specifications concerns WPS application profiles, deployment profiles and transactions, quality in WPS, and grid computing for services demanding large computing resource. Use of geo-spatial ontologies is found in many projects.
Others make use of Web 2.0 methodology, for example to capture human response to flood-loss estimation or to define elusive boundaries, for example the precise area of a city centre. Some papers discussed practical problems in market sectors such as tourism, where mobile devices (PDA) are used to assist people in moving around. An interesting paper from Queen’s University (UK) presented personalised and spatially aware maps, an area as yet little researched.
Three keynote speeches were given. Ismail Chivite, ESRI Redland, discussed how ArcGIS Server fully supported the web, as well as standards such as SQL, OGC, KML, aimed at simplifying the life of a user and developer. The starting point for Dan Cornford, Neural Computing Research Group, Aston University was that real-world problems such as climate modelling and environmental monitoring often required the integration of a range of different observations, models and methods. Not only observations, he said, but also models were affected by uncertainty. Automation did not reduce, but rather even increased uncertainty. He went on to demonstrate how uncertainty could be formalised in UML-like notation.
OGC discussed the benefits of open standards and OGC membership and announced the next OGC interoperability experiment, to take place later this year.