Participation – ESRI has been a GEON partner since the start of the project. There are some similarities in concepts embodied in ESRI’s Geography Newtork (www.geographynetwork.com) and ArcWeb services (http://www.esri.com/software/arcwebservices/), and some of the goals of the GEON portal. Thus, there is natural synergy between these efforts. More specifically, ESRI will make available data sets of interest to GEON, from the Geography Network, via a GEON node at ESRI, Redlands. Authenticated GEON users will be able to access this content as well as all the ArcWeb services. ESRI is also collaborating with GEON/SDSC on ontology-based technologies for data registration and search. Furthermore, we are working towards developing a link from the GEON portal so that scientists who generate results in GEON can easily publish their data and results in the Geography Network and/or using ArcWeb services.
Project Participants
Georeference Online provides in-depth expertise in semantic integration and ontologies for earth science data. Clinton Smythe participates in GEON meetings and provides valuable input in terms of possible industry applications of the use of advanced knowledge representation schemes for earth science data.
Project Participants
Participation – We have been working with Inderpal Narang (IBM Distinguished Engineer), Vithal Gogate, and their team on data grid technologies. In particular, we have been investigating the use of IBM’s GMR (Grid Movement and Replication) software for replication of files and databases in a Grid environment. For related articles, see Services for Data Access and Processing on Grids (http://www.cs.man.ac.uk/grid-db/papers/DAIS_DataServices.pdf) and Towards an Information Infrastructure for the Grid (http://www.research.ibm.com/journal/sj/434/bourbonnais.html).
Project Participants
Participation – HP has provided a 9-node Linux cluster which is being used to power some of the Web services and other applications invoked from the GEON portal. The system consists of dual-processor nodes, with 2GB memory per node. One of the nodes, however, has been configured with 24GB memory to provide support for data intensive computations.