title: "Inaugural Workshop of the ERCIM working group on Semantic Web" date: 27th, 28th May 2004 location: Crete, Greece author: Joost date: 2-6-2004 http://www.ics.forth.gr/ercim_meetings/ http://www.cordis.lu/ist/workprogramme/wp0506-consultation.htm Together with the 15 anniversary of ERCIM this workshop was organized to see whether there was enough common interest in forming an ERCIM work group. The first day was intended to get to know each others work. There were about 20-30 people in the audience of which 7 presented a full paper (45 min presentation) and 4 position papers (20 min). The organizers Dimitri Plexousakis (FORTH), Mark Roantree (DCU) and Thibaud Latour (CRP) had expected more people to attend but this, they thought, was due to overlap with other conferences. My general impression was that there were a few logicians who gave interesting talks, but from a theoretical perspective. In contrast, there were not that many application oriented papers which made the discussions a bit unbalanced initially. This however might be true I think for such semantic web meetings in general, were theorist are more of a unity then the pragmatists. Although I could recognize some of the problems discussed I think the work group will be to small for us to be of real interest. Nicolas Guarino: "The role of foundational ontologies for the Semantic Web." The talk stressed the importance of philosophy and literature and the history they have in knowledge modeling which should not be forgotten during development of the semantic web. Nicolas was especially interested in "capturing" or "grounding" meaning which was necessary for machines to understand. Initially I thought he was talking complete nonsense, and I still think most of it does not make sense, but after sharing a table during lunch I understand him a bit better and what he really was after was ontology evaluation: If you buy an ontology how do you decide it is suited for your domain? Personally I think you model your domain with the use of an ontology and not the other way around, evaluating an ontology from that perspective does not make sense. Harald Gall: "Towards Semantic Web Engineering: WEESA - mapping XML Schema to Ontologies". This talk was about a framework to automatically annotate an XML document with it's structure derived from an XML Schema in order to prevent storage of redundant information. For example a person in XML schema is defined to have a name, the RDF annotation also describes a person as having a name. If a person gets an address as well the RDF annotation is automatically updated. I did not read the paper but it might be a ref if we are including RDF within our generated presentations. Joost Geurts: "Video on the semantic web, experiences with MS." I think the talk was received well. Most of the questions were about video manipulation though and not so much semantic web related. One remarkable question was if I thought about the ethics of this work, since the technology might be used to generate news or documentaries as well while this might not be what really happened. They were a bit shocked when I told them about Stefano's work generating biased documentaries automatically. Noel King: "Service for large scale p2p networks" was about querying p2p networks George Kokkinidis: "Semantic Query Routing and Processing in p2p database systems: The ics-forth SQPeer middleware". Grigoris Antoniou: "A system for Non-monotonic Rules on the web" Was about a system which could reason with imperfect information Peter Vojtas: "Aggregation functions for semantic web" Was about query imperfect information. These were the full papers. There were also a few position papers, some of which were utter nonsense: "In this short communication, we shall try to present the ideas we are currently developing to set up a generic semantic web processor capable of assimilating any ontology built upon any language.". There was one digital library person talking about georeferenced annotations and the problems with it. Pascal van Eck from Twente talked about requirement engineering for web services. The next day, Friday was about the setup of a working group. The objective of the WG was to provide an ERCIM counter weight to the current big European players. Currently there is interest in joining by the following institutes: FORTH (Greece) chair CNR (Italy) w3c coordination VTT (Finland) industry coordination DCU (Ireland) INR (Luxembourg) Workshop organization SICS (Sweden) Workshop organization CSTAKI (Hungary) Workshop organization There was discussion about getting funding. Potential candidates were SWIS and COST. A joint proposal for these will be written and discussed in the next meeting held in Prague august 19-20. If we want to join we have to mail dp@ics.forth.gr. In my opinion we should not. The group is too small to form more the 2 maybe 3 clusters and I don't think there will be a topic which particular interest us. Nevertheless we can wait and see how the proposal looks like and join in if we see collaboration opportunity.