SOCoP Meeting Minutes from Wednesday SOCoP Meeting Sept. 22 2010 from 11:00 - 12:00 EDT (2IH1)
Attendees - Participants Gary Berg-Cross (Knowledge Strategies), Peter Yim (CIM3), Mike Dean (BBN Technologies), Nancy Wiegand (U of Wisconsin ), Krzysztof Janowicz (Penn State), James Wilson (JMU), Dalia Vernanka (USGS) (2IH2)
Following introductions the following were discussed: (2IH3)
1. Workplan for 2010 including a. Update on the ongoing Demo development. (2IH4)
While Todd could not participate he provided 3 slides as a summary of the demos progress. The SameAs service complete enough to use. Here is how it works. One can submit a URI and the service will return owl:sameAs triples for the given URI. The demo also supports a placename search that will return any URIs that has the given placename attached to it (rdfs:label). The other main query operation is a bounding box search that returns any URIs in a given Geo RSS Box submitted by a client. (2IH5)
This search service still needs to be populated with some initial data (as discussed in the previous meeting in concert with available USGS data). When populated it will then plug easily into the Marble Linked Data browser service which is what the user will see and query. Todd estimates that he probably has 1-2 weeks of loose end coding, fixes and data population and the be ready and hosted on the web. (2IH6)
b. Update on USGS integrated National Map (TNM) (2IH7)
The URL to some of TNM data ( no account needed) is: http://usgs-ybother.srv.mst.edu (2IH8)
Nancy and her students looked at some of the data and some representations for bounded boxes and also at vocabulary defs from VoCamp. This raised in her mind the question of whether we want to get an agreement on these spatial representation definitions. Is it too big or too small and issue and do we want to get involved? (2IH9)
Krzysztof reported that Todd and he will try to do a second VoCamp focusing on a top-level for geographic feature types. This should not be a heavy-weight ontology but a collection of ontology design patterns. In the W3C Semantic Sensor Network Incubator group we have just developed such a pattern for sensors and observations. The ontologies should be integrated towards a 'VGI/ citizens as sensors' ontology. We hope to run the VoCamp during the ISWC 2010 and have another one in NY next year. In related discussion it is possible that we could run a VoCamp at the AAG. Todd may also have something to say on these issues. (2IHA)
In a possibly related activity Dalia noted that she wanted to develop an Ontology Users group. (2IHB)
c. Fall Workshop We are still looking for a site since the MITRE space is occupied in OCT and DEC. One possibility was to use the Regional USGS in Reston VA. We discussed a variety of dates in November and there were conflicts with various meetings (e.g. Auto-Carto and USGS meetings). After some discussion following the meeting Friday Dec. 3rd looked like the first good opportunity and Dalia has put in a request for a conference room in Reston. 2. Update on the NsfCybernfrastructure grant award - Nancy (2IHC)
Wiegand (UW) and plans for starting work. (2IHD)
The grant has been awarded but is still working its way through the UW grant process Still hassling with getting the NSF grant set up. Discussion at the September meeting followed from Nancys earlier thought that it was best to regroup and use the time to update our thinking and reach out to communities so we would be ready for when they would meet. One idea is to have a workshop preceding the UCGIS Winter Meeting being held in Washington DC Feb 3 -4, 2011 (see http://www.ucgis.org/). Jack Sanders is the executive but no one has spoken to him yet. Kathy Hornsby is chair of sub-committee and would be interested in spatial ontologies. UCGIS is looking for use cases and problems that people are struggling with for semantic interoperability. Nancy was not sure if we would use our money or if people would come to the workshop on their own. We may be able to rely on that audience for interest and not have to pay to have them at a workshop. Dalia thought that the cost of UCGIS might not be that expensive for participants. Especially if we just do a half day. The group agreed that the timing and location is good, and we did have this on our list in the proposal. The list shown below is larger than we proposed but was consider in an early draft which had more venues: Year 1 (2IHE)
3 trips of 2 days each to DC (NSF, USGIF Tech Days, UCGIS?) 1 trip of 2 days to San Diego (ESRI UC) (2IHF)
Year 2 (2IHG)
2 trips of 2 days each to DC (USGIF Tech Days, UCGIS?) 1 trip of 2 days to San Diego (ESRI UC) Year 3 3 trips of 2 days each to DC (USGIF Tech Days, UCGIS?, NSF) 1 trip of 2 days to San Diego (ESRI UC) (2IHH)
We might target them as an informational session to publicize our work. We should think about whether we would just do an educational session or if we're strategically looking for a domain and use case. One issue with USGIS is that membership and voting is limited to 2 members, so it is a smaller audience than Association of American Geographers (AAG). Their spring (April 12-16) meeting in Seattle and it is a large conference (~7000). There are geographers of all kinds there and it is international. AAG would be a great networking opportunity will include some discussion of temporal issues as it relates to GI Science: See http://www.aag.org/cs/giscienceresearch As mentioned in the August meeting the deadline for abstracts is in Oct. 20th. In a session we might publicize what we are doing. And we might call for participants on the topic problems to be solved or what they are working on. Nancy thought that we might not be able to do both USGIF and AAG. She thought that AAG would be a better opportunity than USGIS. Dalia was planning on attending AAG and would love to have someone to work with. James added that our topic would be of interest to AAG folks. Would people have to pay to attend a workshop? (2IHI)
Other conferences discussed included: (2IHJ)
ESRI Federal Users conf in January 2011 (2IHK)
2011 NSGIC Midyear Conference February 27 - March 2, 2011 Annapolis, Maryland (2IHL)
Another thought is that we might do one with USGIS folks and another with Federal folks who are available in the DC area. (2IHM)
There is a Summer UCGIS opportunity in June 23rd in Bolder. (2IHN)
See http://www.urisa.org/conferences/calendar for a calendar of some relevant conferences. (2IHO)
There was a small discussion on setting up Cyberinfrastructure (2IHP)
Gary and Peter had arranged to talk through some of the infrastructure issues on Sept. 23rd. One question was will we have a separate server for INTEROP? In response Peter noted that the Collaborative Work Environment (CWE) for INTEROP is already collocated with Ontolog. The infrastructure has already been set up and we can start using the (webdav-based) shared-file repositories any time now. Currently these are not populated. Gary needs to organize and manage the content. The site is at www.Socop.org (2IHQ)
We do have a Forum set up for SOCoP, but we might want to consider 2 additional ones. One for the core team and another for the INTEROP members, which has partial overlap with SOCoP. The group agreed to forming an INTEROP mail group but be inclusive in it. (2IHR)
Another issues was the infrastructure for virtual events like OOR incubator and the 2009 workshop which captured the talks and slides and also some notes from the meeting. (2IHS)
Gary noted that the CWE / Cyberinfrastructure should support (i.e. access, integrate, automate, and manage complex, collaborative projects across disciplinary as well as geographical boundaries) INTEROP members not just him. A target is to be user friendly enough for members to easily add content. It needs to be easier and Gary took an action to write it up some of the subsequent discussion of this and include as part of the meeting. A paragraph on this is attached at the bottom of these minutes. (2IHT)
Another consideration is to have a seamless connect to the OOR. (2IHU)
As noted previously Nancy had looked into using Hubzero as part of the cyberinfrastructure (see https://hubzero.org/) but has not created an instance at Wisconsin. If this is available at no cost, we might be able to use this as a supplement and point to it for some things. Peter noted that CWE is designed to be totally opened to link to other sites. (2IHV)
3. Other Business (2IHW)
Krzysztof also summarized the Linked Spatiotemporal Data workshop at GIScience. (2IHX)
There were about 28 participants and a lot of interesting papers, live demonstrations, as well as lightning talks. The papers are available online at http://stko.psu.edu/lstd2010/ . There will be also an *open-call* special issue in the Semantic Web Journal (http://www.semantic-web-journal.net/) about Linked Spatiotemporal Data and Geo-Ontologies with a deadline in mid of December 2010. At the workshop participants also discussed the upcoming research challenges during the workshop and two directions seem to be important for most people: (2IHY)
- work on GeoSPARQL and - work on vocabularies/ontologies. (2IHZ)
Moreover, most attendees seem to agree that encoding GML:Geometry in RDF is probably not a good idea as it does not add any semantics, retrieval, or reasoning capabilities but leads to a huge overhead (and will be very slow). Sven Schade and Michael Lutz had more examples about what should be transformed to RDF and what should not. For them the important part is how to link the RDF data with legacy data and other information that will not be available as linked data. There seems to be a need for a Linked Data Model for the Geo-sciences in addition to conceptual models and data models. One crucial aspects of such a model would be outgoing and incoming links. We had some interesting discussions on this during the workshop and the GIScience. In terms of Sensor Web Enablement we already started to do some initial work on this; see http://www.personal.psu.edu/kuj13/DE2010lsd.pdf. (2II0)
There were also some interesting discussions on the semantic enablement of existing OGC services, semantic annotations, as well as light-weight (micro) SDIs during the GIScience 2010. There is also some new work on Linked Data Services. (2II1)
Krzysztof noted that there will be a workshop on OGC's SWE and the Semantic Sensor Net at the ISWC 2010. (2II2)
Dalia on RDF MD issue Dalia asked about best practices for documenting triples something to replace things like the old FGDC standards and adapt then to semantic tech. Dalia would welcome any thoughts on this. Krzysztof provided some ideas from the Linked Spatiotemporal Data workshop at GIScience. Mike suggested ideas from the Vocabulary for interlinked data (VOID). John suggested bouncing the idea off of Sharon Shin, Dave Danko and Jean Brodeur. The group agreed that this night be a good topic for next year or the workshop. (2II3)
There being no other Items or New Business the group set a tentative date for the next planned meting as Wed Oct 20th 2010. (2II4)
Notes on the Sept. 23rd CWE meeting (Peter and Gary) (2II5)
The SOCoP wiki can readily add text, for those unfamiliar with editing the particular flavor of wiki, some training (self-training by reading up the documentation, orsome group training to be organized) would be necessary.. There is a online page to understanding this- Ontology wiki Basic Text Formatting. There is also an Ontology Style Guide. These may not be intuitive to some users and will have to be studied. Peter thought this a small investment. The CWE is converting over to MediaWiki and this would work towards a better UI but is not a 1-2 week thing. (2II6)
In addition some special utilities are needed to add files. The CWE uses webdev for adding files. Peter took Gary through the process for loading a free DataFreeway for a VISTA OS. Once it is installed and points to the proper Ontolog folders one can drag and drop files to the webdev folder first and them can point to it on Wiki pages. Again this will need some time for INTEROP members to master. Gary is set up to do this, but it might take some time for members to be set up and facile. Peter suggested that those who plan to actively use the CWE to go through the Ontolog-WikiHomePage (especially the sections on "Ontolog Community Wiki" (especially the paragraph at: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nidA). Peter also offer to conduct a virtual training session, and that Gary might want to organize it if enough people are interested to participate. (2II7)