See also: IRC log
https://www.w3.org/2016/10/03-poe-minutes.html
RESOLUTION: the minutes of the last call are accepted
<phila> michaelS: No, I haven't
Simon was active on adding references
renato: key issue is that we
currently have only a single constraint, how to extend
this?
... One of the use cases were in Lisbon: a relative constraint
like "may be published 30 minutes after the game"
<benws2> +1
phila: doesn't seem to me so over-complicated
<Brian_Ulicny> +1
phila: supports to include it.
ivan: is this about only a time
constraint - or a more generic requirement
... But this could be solve by starting with the time
constraint first and then to make it more generic
ben: thinks this is a general
issue, not only about time
... there are use cases related to payments which are very
similiar
... we need a generic solution
renato: we need also to express relations between constraints - something "constraint A depends on constraint B"
phila: are all relative time use
cases about a related event
... in this case we only need a new "related event"
<phila> ac me
ben: another relative relationship is : payment per account, payment per person
ivan: asked renato to put in IRC his draft of a discussed new solution
renato: has no clear outline,
only first ideas
... we need some thoughts how to do that in the ontology first,
then think about encoding
... issue: we have to be aware that one thing may occur before
the other ... maybe next week.
ivan: should we reorder the sequence of solving things
renato: ok, we could look at relatives time constraints first
<James> sounds a good approach to try
<Sabrina> Seems like a reasonable suggestion to me!
phila: use case "30 minutes after game ended" - what is the right expression to express this logic
ben: maybe the best way to test this examples
renato: constraints on
constraints or dependancies across constraints?
... let's talk about extended relations
... to express alternative options, like 'pay' or 'subscribe to
a service'
... what is the model for hat?
... does W3C has similar solutions for tthat?
ivan: cannot recall such a
solution. What renato describes is very close to the previous
use case
... - first you have the triples of the statement and then you
want to make a statement about one or more of these
triples.
... this is the typical case for using a named graph
Brian_Ulicny and ivan discussed options for solving that issue
ivan: thinks what is discussed must be expressed properly in RDF
ben: is this the case for latests suggestions?
<Sabrina> Agreed.... This is what we need to look into....
ivan: not sure, needs to be checked in detail.
<Sabrina> I"m happy to look into this
<Sabrina> q
phila: feels this goes in a very
complex direction
... we might go outside OWL - will users want to do that
Discussion about the need for a reasoner.
renato: we need to make a model and include that into the existing one but next we need to fit this into the OWL of the ontology
ivan: we might define a
reifiction-like object which might have additional information
attached to it
... we might apply a specific URI to a constraint
... in a next step we could created a list of URIs/constraints
and apply a next level of relationships/dependancies to
them
... we should try that - may there are hidden problems
renato: the additional
constraints should be added as additional data - the basic date
should always be the same
... it would be great if some come up with example solutions
this week and we could look at them at the next call
<James> I don't mind
renato: who volunteers for that`?
<James> sure
<benws2> Me too
<James> Be good to have more than pair of eyes
<James> *one pair
ivan: voluteers to look at the examples, but will be in China next week
<Sabrina> I will look into OWL
<Sabrina> I am also not here next week as I will be at ISWC in Japan
renato: will communicate next steps with the volunteers
<victor> (please count me in)
<Brian_Ulicny> https://en.wikipedia.org/wiki/Allen%27s_interval_algebra
<Zakim> phila, you wanted to talk about Allen's algebra
Brian_Ulicny: we should investigate the different types of temporal relations
phila: Special Data WG is behind registering such temporal relations
renato: how to express that a policy is only valid in a stricly defined period - by what means to express that?
<phila> LOV search for valid from
ben: aren't there already existing time constraint properties?
phila: searched for "valid from" and found many
ivan: looks a provenance ontology, has some very specific time related properties - this group did not try to find a generic solution
<James> I think we want "expires"
renato: common wording is "valid
from" and valid to"
... clarified the suggestion of "expires": is the same as
"valid to"
victor: how to proceed, by exchanging emails?
renato: yea
renato: the group has discussed already what should be put on this list. There is a list of documents in the Charter
<renato> https://www.w3.org/2016/09/poe6
renato: went over the items shown in https://www.w3.org/2016/09/poe6
phila: will support finding a solution for item 5, mapping
<renato> https://www.w3.org/2016/09/poe8
renato: went over the draft at https://www.w3.org/2016/09/poe6
victor: volunteered to organize the f2f meeting in March - details at the next call
ivan: will send out an email about what he discussed above
renato: thanked all participants and closed the call