See also: IRC log
<KJanowic> Hi
<ahaller2> \me hi
<kerry> trackbot, start meeting
<trackbot> Meeting: Spatial Data on the Web Working Group Teleconference
<trackbot> Date: 09 August 2016
<kerry> scribe: roba
<kerry> scribeNick: roba
<KJanowic> Hi
<kerry> https://www.w3.org/2015/spatial/wiki/Meetings:SSN-Telecon20160809
<kerry> http://www.w3.org/2016/07/26-sdwssn-minutes
<kerry> +1
<ClausStadler> +1
<ahaller2> +1
<SimonCox> +1
+1
<JRamsay> +1
<KJanowic> +1
RESOLUTION: approve minutes http://www.w3.org/2016/07/26-sdwssn-minutes
<KJanowic> Kerry, I think you are turning your head away from the microphone
<KJanowic> you can also find examples at: https://github.com/w3c/sdw/blob/gh-pages/ssn/rdf/documentation_examples/sosa-core_examples.ttl
<SefkiKolozali_UniS> +present SefkiKolozali_UniS
<kerry> KJanowic: sosa core is what is the smallest set of terms that can be used to annotate bilions of observations on the web
<kerry> ...e.g what does the av web dev need as a minimal set to post observations online?
<kerry> ...also with semantic hooks for other more complex app areas
<kerry> ...an ontology design pattern
<kerry> ...many different aplications will be based on core and various applications based on core can be compared by sosa terms
<SimonCox> SOSA can stand on its own for lightweight applications, large numbers
<kerry> ...it stands on its own as a lightweight ontology for billions of obs and also a common core for many more complex applications
KJanowic: SOSA should stand on its own as a lightweight vocabulary, but also be common core
<SimonCox> i.e. also basis for richer semantics
KJanowic: reminds us its not a physical data model :-)
<SimonCox> OWA - features not present in SOSA-core does not mean they are out of scope
<KJanowic> https://github.com/w3c/sdw/blob/gh-pages/ssn/rdf/sosa.ttl
ack \
scribe: observableProperty of a
featureOfInterest is what a sensor result refers to
... needs to support _repeatability_ of observations
... Procedure is a recipe reference that can be looked up
... superclass of Result - any data type
<KJanowic> https://github.com/w3c/sdw/blob/gh-pages/ssn/rdf/documentation_examples/sosa-core_examples.ttl
<KJanowic> SELECT ?observation WHERE { ?platform a sosa-core:Platform ; ^sosa-core:hostedBy ?sensor. ?sensor sosa-core:madeObservation ?observation. }
<kerry> roba: sosa core describes things as owl classess and object properties where I thought it should be rdf with owl making those assertions in a spearte model
<kerry> ...also examples was looking for relation between obs , sample, and foi, but sample is not referenced in example so cannot see how to get to foi
<kerry> KJanowic: seel line 39
<kerry> roba: yes but cannot see how to get from obs to that
<kerry> KJanowic: yes, not there, just did not get to it
<kerry> robs: rdf scope/
<kerry> KJanowic: not about rdf vs owl -- about what is intuitively understandable without over burden, a little bit of rdfs, a little bit of owl, a litt bit of scheam.org
<KJanowic> @Roba: this is not about rdf vs owl but about a fragment of Web-scale KRs that is intuitively understandable to Web developers, end users, and so forth.
DanhLePhuoc: uses syntax from owl that implies logic - what is intent here?
KJanowic: uses meta - annotates
without introducing owl logic
... inverse is useful to help Web developers formulate the
right queries
kerry: important issues are discussed on public email list - github tracker not expected to be visible to all
<SimonCox> https://github.com/w3c/sdw/issues/336
<KJanowic> https://github.com/w3c/sdw/issues/334
<SimonCox> https://github.com/w3c/sdw/issues/309
KJanowic: issues posted to public list when discussion settles down - commits and comments on merges
<KJanowic> https://github.com/w3c/sdw/issues
<SimonCox> For discussion on inverse properties - see https://github.com/w3c/sdw/issues/334
<KJanowic> @roba: yes, exactly. Back and forth on specific lines of code on github, bigger picture and regular updates on the mailinglist
kerry: PROV removed inverse
properties - to stay within OWL-RL - noted in comments
still
... introduces heavy demands
... asks to revisit
<KJanowic> inverseOf open for discussion.
DanhLePhuoc: back to logical
impications
... inverse requires entailment in store to be used (did i get
that right?)
<kerry> +1 to danh's concern
KJanowic: most high performance 3-stores dont use any rules (is sad... but accepting)
DanhLePhuoc: notes at least RDFS reasoning supported
<ahaller2> ahaller2: I was agnostic, but a bit sceptical about the inverse properties in the core, and it seems the majority here does not want them in the core
KJanowic: allows data to be compressed
<kerry> +1!!!!
<KJanowic> Will do
<KJanowic> I promise
<SimonCox> Yes - we should put a stub email on the list with link to issue in github
<kerry> roba: wants to see core linked to owl-dl entailment (not sure I got this) butannotation
<kerry> ... would be happy to live with rdfs but want to see the owl-dl as well
<kerry> +1 to rob's comment
<DanhLePhuoc> agree with Jano that : triple store can specify the entailments they could support: https://www.w3.org/TR/sparql11-entailment/
<DanhLePhuoc> +q
<kerry> skos does have formal semantics as defined by owl-dl -- just "simpler" semantics
<kerry> issue: what reasoning language should the core use?
<trackbot> Created ISSUE-72 - What reasoning language should the core use?. Please complete additional details at <http://www.w3.org/2015/spatial/track/issues/72/edit>.
<KJanowic> Thanks Roba, Danh, this is super, super helpful!
SimonCox: audience is key - been thinking about the Web Developer use case - not currently looking at others
<ClausStadler> https://www.w3.org/TR/vocab-dcat/#vocabulary-overview
<KJanowic> check also the "semantic web for the working ontologist" book by Jim and Dean on inverse properties
ClausStadler: inverse means that can make relationship work from the related resource - dont have to update the source
<KJanowic> +1 to Claus
ClausStadler: but useful to have inverse properties in spec so dont have to hijack the ontology namespace
why cant we formalise in sosa-core-owl.ttl ?
<kerry> ack
and have inverse in core, without OWL axiom (annotation) and achieve this?
DanhLePhuoc: why not describe inverse as descriptive property
KJanowic: needed inverse property in SPARQL
isnt that also a "choice of the triple store?"
<ahaller2> @roba we will formalise the core in the vertical layers, if we decide that there is only RDF(S) in the core, it would be formalised in the vertical core, so that would fulfil your requirement
roba: SPARQL Use Case is important - but introduces an interoperability contract between the querier, the triple store's entailment regime and the underlying ontology
<KJanowic> @all: Thanks a lot! This was very helpful and productive.
<kerry> https://www.w3.org/2015/spatial/wiki/Attending_F2F4
<KJanowic> @roba: I have to think about this. I dont have a ready answer right now
roba: asks for the audience and such contracts to be much more clearly expressed in documentation (included embedded documents)
<ahaller2> @KJanowic you are not coming?
kerry: asks for potential attendees at F2F to identify themselves
<KJanowic> @ahaller2: most likely not, we are expecting this fall
<ahaller2> @KJanowic fair enough, that's an excuse :-)
<SefkiKolozali_UniS> I can participate online
<kerry> https://www.w3.org/2015/spatial/track/
<KJanowic> Thanks everybody!
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/..also/...also/ Succeeded: s/..../.../ Succeeded: s/tracker/github/ Succeeded: s/bu /but/ Succeeded: s/identy/identify/ Found Scribe: roba Inferring ScribeNick: roba Found ScribeNick: roba Present: kerry KJanowic ahaller2 ClausStadler JRamsay ScottSimmons SimonCox roba DanhLePhuoc SefkiKolozali_UniS Regrets: phila Found Date: 09 Aug 2016 Guessing minutes URL: http://www.w3.org/2016/08/09-sdwssn-minutes.html People with action items:[End of scribe.perl diagnostic output]