Minutes of the February 2010 SOCoP Meeting    (2IXG)

Wednesday Feb. 17, 2010, from 11:00 - 12:10 EDT    (2IXH)

Participants: Eric Wolf (USGS), Dalia Vernaka (USGS), Gary Berg-Cross (Knowledge Strategies), John Moeller (NG), Peter Yim (CIM3), Todd Pehle (Orbis Technology) , Lynn Usery (USGS)    (2IXI)

Over time some additional links may be added to related material and sections on topics created that accumulate and organize material from successive meetings.    (2IXJ)

The agenda included 9 main topics:    (2IXK)

1.Update on USGS integrated National Map (TNM) data 2.SOCoP Demonstration and Development of GeoSpatial Service 3.GUI Tool selection and use 4.Ontolog Announcements 5.Update on OGC RFC specification (OGC 09-157) 6.Open Ontology Repository efforts and related SOCoP activities 7.NSF INTEROP proposal 8.Update on future use of Ontolog SOCoP Forum for discussions. 9.CFP – 4th International Workshop on Semantic and Conceptual Issues in Geographic Information Systems    (2IXL)

Summary of Topical Discussion    (2IXM)

1. Update on USGS integrated National Map (TNM) data becoming available    (2IXN)

The National Map has multiple data layers, including environmental and human use information, on top of the base map that goes beyond other mapping efforts. There is new, integrated, hydrology data is stored in Oracle and represented in owl. This set of data for 6 watersheds and 2 urban areas (St. Louis and Atlanta) will include all 8 data layers and become available for the watershed area.    (2IXO)

Eric Wolf, a USGS employee, coordinating work with Dalia Veranka, described the progress made toward having National Map data available in triples for the St Louis watershed. He put up N3 files last week . Eric also wrote a Python script to generate a linkage to Oracle. What is there is just attribute data in N3 and almost complete on the geometry (was in KML) but changed to GML per recommendations from OGC. He expects to generate RDF soon.    (2IXP)

Dalia noted that this is a new accomplishment and exciting. Dalia also noted that metadata has not yet been developed for the updated data. Eric and Dalia will have documentation for the public, for example, examples of the landset types. This requires updating the information from the older Shape file info on the different landscape types (urban, rural etc.). Gary might help with this documentation update.    (2IXQ)

Lynn Usery discussed the set up of a 3 TB Server which will be outward facing and web service at Missouri University of Science and Technology.    (2IXR)

This server could be used for hosting a service. USGS needs to set the rules for use of the server, for example on the use of distributed data and other services such as proposed by Todd. One issue raised was the use of Oracle Spatial. USGS has enterprise license for Oracle Spatial and will investigate enterprise licensing issues. Dalia noted that USGS has other sites where they host with this type of functionality.    (2IXS)

Todd will also check with Xavier Lopez on possible use of Oracle Spatial.    (2IXT)

2. SOCoP Demonstration and Development of GeoSpatial Service    (2IXU)

In the context of an overall demonstration, Todd Pehle updated the group on his work to implement a “geo/spatial SPARQL” query ability in the SAIL (Storage And Inference Layer) . Todd is starting to build a Sesame Sail on top of Oracle Spatial. It will use some SPARQL operators (service endpoint) to showcase the concept working in coordination with the OGC spec.    (2IXV)

For example we can use it to calculate the distance between two points. We might do what Marco Neumann and Taylor Cowan did with their geosparql tool built on Google’s new app engine for java which extends Jena. This engine can process simple polygon searches for the concept of “near”. See http://geosparql.appspot.com/ for an online service. Todd notes that it uses an in-memory model so it may not scale to millions of triples, but Todd felt sure we could still make use of it on smaller datasets, such as we would start with. That is we could set up few smaller datasets and host these on Google appspot (e.g. http://appgallery.appspot.com/) just like Taylor did. We could then use it as another SPARQL endpoint in our demo. (Simple a SPARQL endpoint is located by a URL on the Web that implements the SPARQL protocol. Generally this means that the URL can be sent queries.)    (2IXW)

Todd and Eric have been in discussion on coordinating their efforts. Todd will meet with Eric while in Denver next week so they can mature these idea. They might work some examples that will moving this along from a relatively abstract use case to something more concrete. Part of this discussion might be about how to use sparql end points and rdf data coming back as query results. The use case might need another service to make it more “useful”, processing it into easier for novice users to understand etc. Under some scenarios we might use Geonames (http://www.geonames.org/ a geographical DB, covering all countries, holding over eight million place names for things like cities, villages, lakes, parks, hotels etc.), Open Street Maps and use Google Maps for images. (see January minutes for a discussion of an effort by USGS to leverage such vocabularies). An example of a simple demonstration using these is at http://3liz.org/geolocation/    (2IXX)

Another possibility is do a Mashup with DBpedia data and existing endpoint. There are other existing endpoints we can use. Todd noted that he was tracking some relevant semantic web projects using OGC spec. One at Wright State and another in Australia.    (2IXY)

Todd noted that he has some “sped dial” numbers for some people, who may give advice and help with resources. We should also check with Dave Kolas who has implemented some parts of this. Also Franz Inc. might be interested in participating to showcase their tools.    (2IXZ)

One question was when we would demonstrate such a service. Gary noted that we generally have 2 phases. The first targets functionality for the May or June timeframe of the NGA Tech Days/USGITA workshop. John Moeller provided the following details on this year’s workshop. The 2nd phase would be at our Fall workshop (Oct. or Nov.).    (2IY0)

John Moeller noted that he had spoke with Mark Reichardt on how to tie any SOCoP work together. Mark suggested working with the GeoSemantics WG. The next OGC Technical Committee meeting will be held March 9th during th OGC session 8-12 in Frascati, Italy, and Josh Lieberman would be the person to contact. Gary took an action to contact Josh about the WG meeting since there is a call out for topical suggestions.    (2IY1)

We could also work with Xavier on the SPARQL spec to mature our ideas.    (2IY2)

3. GUI Tool    (2IY3)

As part of developing a usable service Todd proposed OpenLayers as a possible tool for Developing a GUI client. OpenLayers is a lightweight, open source, javascript-based “slippy maps” . It can connect to OGC-based WMS services as well as non-standard map services like Google maps or Bing, etc. Gary asked who would use this tool and whether it would be used on the USGS server for development or elsewhere and what “users” would need on there end.    (2IY4)

4. Ontolog Community Announcements    (2IY5)

Peter Yim summarized 2 Ontolog community announcements    (2IY6)

The first, which some have seen, concerns the OpenOntologyRepository (OOR). The OOR was co-convened in 2008 and had a webinar Friday Feb. 19th on SW development with 4 panelists chaired by Mike Dean. (For details on who contributed see the session page OOR / ConferenceCall_2010_02_19)    (2IY7)

Three more OOR sessions planned. On March 25 there will be one Content and on April 1, one on Use Cases (chair is Todd Schneider ). A 3rd session will be on IPR.    (2IY8)

SOCoP is looked on as a group that could contribute to OOR code as well as content.    (2IY9)

The 2nd announcement concerned the OntologySummit2010 that is ongoing. The 3-months of Summit activities will culminate in a 2-day symposium to be held at NIST on March 15-16 as part of 2010 NIST Interoperability Week. See: OntologySummit2010_Symposium    (2IYA)

Registration will open very soon. The theme is "Creating the Ontologists of the Future" and Peter noted that surveys are out to collect representative info on this topic to advance this meeting. See: OntologySummit2010_Survey.    (2IYB)

5. Update on OGC RFC specification (OGC 09-157)    (2IYC)

Due to another set of meetings, Xavier Lopez was unable to attend this morning's SOCoP call. However, there are a couple of items he wrote up to share with the SOCoP group:    (2IYD)

a) Technical input to the OGC RFC are still being considered. Members can email comments, questions, suggestions to me (xavier.lopez@oracle.com)    (2IYE)

b) We are considering members to be formally acknowledged on the spec submission team. Anyone who has contributed (or planning to contribute), AND whose organization is a member of OGC, is welcome. Those interested can contact me.    (2IYF)

Carl Reed provided an email (see below) the outlines/describes the standard OGC spec submission process quite well. In this context it is fair to say that we are barely at the beginning of the specification process.    (2IYG)

Carl Reed clarification of the OGC RFC process.    (2IYH)

The first point is that, Oracle has submitted GeoSparQL into the OGC process. This starts the followng 8 steps:    (2IYI)

1. A Standards Working Group will be formed. To form a SWG, a submission team is formed, a charter written and approved, and a call for participation is done. 2. Once formed, the SWG performs work on the document (if required). 3. When the SWG believes the document is ready, they can vote to release for a 30 day public comment period. 4. At this point, the OGC Architecture Board reviews the document and formally approves (usually with guidance) the release of the RFC 5. A 30 day public comment period occurs. Anyone can submit comments during the comment period. There is an associated press release that announces the comment period and instructions on how to submit comments. 6. The SWG processes the comments and edits the document as appropriate. 7. When the SWG believes the document is ready, the recommend for an adoption vote. 8. An adoption vote occurs. The adoption vote includes a 60 day IPR review.    (2IYJ)

Carl noted that there are other actions that occur during the approval cycle, such as dealing with copyright and intellectual property. However for a well formed candidate standard, this whole process can be done in about 9 months. This is quite quick when considered in terms of other standards organizations. There are also other opportunities for non-OGC members to provide input, such as using the public OGC Change Request app.    (2IYK)

 If anyone in SOCoP  has any questions regarding this process, please contact Carl ("Carl Reed" creed@opengeospatial.org)    (2IYL)

6. Update on Open Ontology Repository efforts and related SOCoP activities    (2IYM)

Gary Berg-Cross noted that he had used the OOR “sandbox” (http://oor-01.cim3.net/ontologies ) and loaded an Ordnance Survey Spatial relations ontology as a start on populating the OOR with relevant geospatial ontologies. Other relevant ontologies there include the Spatial Relations ontology added by Todd Pehle (made up of vocabulary/concepts developed at the NeoGEO VoCamp on Distance, double, Geometry (Box , Circle, Line, Point and Polygon and Spatial-Object) and a larger Spatial Ontology added by Melissa Haendel from anatomical spatial domain (dorsal, axis etc.). Gary noted that there are several Ordnance ontologies and will contact the Ordnance Survey to coordinate on their storage in the OOR. The intent of this work is not only to make a wide range of geospatial ontologies available but to look at their use in support of our demonstrations and geoservices development. Thus over time the OOR ontologies should support be related to such work as that of TNM and vocabulary development. Indeed the National Map watershed model would be a good candidate to load and map to other ontologies stored there. Mike Dean had performed some mapping on the Todd’ spatial relations ontology to GLM’s geometry earlier (e.g. Geometry (Spatial Relations) => gml:_Geometry (GEO_2007) 01/28/10 mdean).    (2IYN)

7. NSF INTEROP proposal    (2IYO)

Via email Nancy Wiegand had noted that here was no change to the status of NSF INTEROP proposal.    (2IYP)

8. Update on future use of Ontolog SOCoP Forum for discussions.    (2IYQ)

Gary discussed his work with Peter Yim (CIM3) to set a a SOCoP Forum for discussion as part of the Ontolog Community. Ontolog is a “an open, international, virtual community of practice devoted to advancing the field of ontology, ontological engineering and semantic technology, and advocating their adoption into mainstream applications and international standards” - see http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage. Ontolog supports Wikis and Forums (see http://ontolog.cim3.net/forum/. ) Having the Forum will allow discussion of topics and perhaps a more consistent relationship with the larger ontology community that discusses these topics and holds regular teleconference meetings with experts in the field. Ontology had participated in our Nov. workshop and recorded the morning session placing it and presentations on their site.    (2IYR)

For the workshop Peter set up SOCoP as a “Project” (the OOR is also a Project) with pages for content. More recently Gary used the current SOCoP mail list to create a subscription to the Forum that is part of being a Project member of Ontolog. SOCoP members should have received a notice of this. People can unsubscribe to the Forum without unsubscribing to the current mail list. Gary asked the Feb. participants if they were interested in the use of such a Forum and Dalia was the first to say they she would like to receive discussions. There being no objects Gary said that following the meeting he send out announcement on how to use the forum and maybe some topics for discussion. One example might be a discussion for ideas on our integrated demo using National Map data. Another topic might be discussion of public data resources for the demo and service as well as candidate ontologies for storing in the OOR. Todd mentioned that he, Christian Berg others were doing a workshop on some of these topics and will have more details on this for the community (perhaps on the Forum).    (2IYS)

The SOCoP forum is listed on the Ontolog Forum page and is at: http://ontolog.cim3.net/forum/socop-forum/ Currently there is only a test message there, but after distribution of the minutes Gary will send an announcement from the Forum to members.    (2IYT)

One guidance from Peter on posting to the Forum is to:    (2IYU)

•“ save the posting address <socop-forum [at] ontolog.cim3.net> to your address book; (29TU) •set your mail client to send & receive in plain text only, and •wrap the text with line length of 65 characters. He asks that poster also, separate paragraphs with a blank line to ensure optimal granular accessibility. This allows proper “purple-numbering” to the archives. Purple-numbers allow allows one to address finer detail (e.g. paragraphs) of HTML documents.    (2IYV)

9. Other Business    (2IYW)

Jean Brodeur sent out information on the Fourth International Workshop on Semantic and Conceptual Issues in Geographic Information Systems ( SeCoGIS 2010) in conjunction with the 29th International Conference on Conceptual Modeling (ER 2010) Vancouver, Canada, November 1-4, 2010. Paper abstract submission are due April 8, 2010.    (2IYX)

There being no Other Items or New Business the next meeting was targeted for March. 17, 2010.    (2IYY)

For more information, contact the SOCoP Executive Secretary - Dr. Gary Berg-Cross via email to: gbergcross [at] gmail.com    (2IYZ)