- From: Luc Moreau <L.Moreau@ecs.soton.ac.uk>
- Date: Fri, 17 Feb 2012 15:04:13 +0000
- To: James Cheney <jcheney@inf.ed.ac.uk>
- CC: Provenance Working Group <public-prov-wg@w3.org>
Hi James, As I indicated in my preamble, I didn't know whether this issue had to be raised against the mapping product or not. I think many of the issues are about the ontology itself. I am afraid I read your email today, well after I raised the issue. My primary aim was timeliness since the owl ontology needs to be fully written before we can talk about alignment in any significant way. With hindsight, this should have been raised against the ontology, which I am doing now. Further comments below. On 02/17/2012 01:43 PM, James Cheney wrote: > Hi Luc, > > Thank you for this excellent non-example of the way I asked for issues against the mapping to be formed. Most of the observations below are vague in the sense that you do not specify whether the observation is a description of a problem or a description of a solution, and none gives much clue whether the change you believe is needed is to the mapping file, DM, ontology, or both. > > According to the process [1] I sketched (to which no one has objected so far), I am going to try to interpret your observations and identify the sub-issues that can be fixed by modifying just one document, and (if you agree) transfer or re-raise them to PROV-DM or PROV-O as appropriate. > > I will try to address any issues that can be fixed by changing the mapping myself or through consultation with the others who've worked on it (Tim and others, please feel free to step in here.) > > Detailed comments inline. Please indicate whether the proposals below would address the corresponding concern or whether I've missed some of your points. > > --James > > [1] http://lists.w3.org/Archives/Public/public-prov-wg/2012Feb/0249.html > > On Feb 16, 2012, at 9:57 PM, Provenance Working Group Issue Tracker wrote: > > >> PROV-ISSUE-253: misc issues with the ontology [mapping prov-dm<-> prov-o] >> >> http://www.w3.org/2011/prov/track/issues/253 >> >> Raised by: Luc Moreau >> On product: mapping prov-dm<-> prov-o >> >> It seems that it's more a fixing bug issue than an alignment problem >> per se, but I raise this issue against the prov-dm<->prov-o mapping. >> >> >> >> >> >> Usage >> --- misses a property hadActivity >> >> > I don't understand what you expect to change here. There is no hadActivity construct in either PROV-O or PROV-DM, so I can't just change the mapping to include this. > > There is a link from the activity to the usage labeled hadQualifiedUsage. The hadActivity link would be the inverse of that. However, the edges here are meant to be consistent with the direction "towards the past". So I don't think we want to replace hadQualifiedUsage with hadActivity going in the reverse direction. > > proposal: Raise against PROV-O to discuss whether to introduce a hadActivity property linking QualifiedInvolvements to Activities. > > > I don't think the ontology consistently follows the "towards the past" approach. It has both generated and wasGeneratedBy (which is ok by me). >> ---------------------------------------------------------------------- >> >> Bundle: not part of DM3? >> > I don't understand what would address this issue. There is a Bundle section that contains some discussion of Account and RecordContainer. Since Account was put on the endangered species list at F2F2, my impression was that they were not required to be handled in the first draft of the mapping. > > Please be more specific about what aspects of bundles you think aren't handled here and should be. > > Proposal: Defer until status of bundles/accounts/record containers is stable. > > We said that we were aiming at alignment with WD3 - Accounts. So, it is out of scope. >> ---------------------------------------------------------------------- >> >> There is no time information associated with Entity in DM3 >> >> > Correct, but I don't see what you think should change (in the mapping, PROV-O, or PROV-DM). The rule for entity() records does not link the entity to a time. It is possible to link any Thing to a time, including an Entity, but so what? > > Proposal: No change. > > It's an example where the mapping is not implemented by the ontology. The issue is inter-operability, really. If no time information is expected there, it shouldn't be allowed. >> ---------------------------------------------------------------------- >> >> >> Association: >> hadQualifiedAssociation property missing >> > I interpret this as an issue with Prov-O, which (if addressed by adding such a property) should be reflected in the mapping. > > Proposal: Raise against PROV-O, and reflect any ensuing change in mapping. > > agreed. >> ---------------------------------------------------------------------- >> >> Association: >> hadQualifiedEntity has range Entity, >> but it should be Agent .... >> hadQualifiedAgent with range Agent, >> >> > This is a PROV-O issue, which if addressed there can be easily reflected in the mapping. > > Proposal: Raise against PROV-O, and reflect any ensuing change in mapping. > > > agreed >> ---------------------------------------------------------------------- >> >> Association >> --- misses a functional property hadActivity >> >> > Same response as for Usage. > > proposal: Raise against PROV-O to discuss whether to introduce a hadActivity property, and reflect any ensuing change in mapping. > > agreed. >> ---------------------------------------------------------------------- >> Association >> ---- adoptedPlan i would have thought it had to be functional >> > This is a PROV-O issue. > > proposal: Raise against PROV-O. No change is needed to the mapping. > > agreed, >> ---------------------------------------------------------------------- >> >> Delegation: what is it? >> is it what is called Responsibility Record in WD3? >> >> > I'm not sure (didn't write this part), but I believe it is a class that is populated by the identifiers of responsibility records (as Usage for used, Generation for wasGeneratedBy). This seems obvious from the way it is used in the rule, but deserves explanation. > > Proposal: Add a sentence to ProvRDF explaining this. > > why aren't we using the same term as in prov-dm? >> ---------------------------------------------------------------------- >> >> No collection >> >> > True, and the reason is the same as for Bundle - constructs that were in-flux or endangered as of F2F2 were not expected to be mapped. > > Proposal: Defer until collections stabilize. > > I didn't think it was the case, but fine, if it's the consensus >> ---------------------------------------------------------------------- >> >> HadTemporalValue >> --- is not functional >> > True, but this is a PROV-O issue. > > Proposal: Re-raise against Prov-O. No change needed to mapping. > > agreed, i think the ontology does not implement the mapping since it allows the range to be interval. >> --- has QualifiedInvolvement in its domain but >> Association and Delegation don't have temporal information >> >> > This is an example where the mapping may suggest changes to PROV-DM. > > It's true that in Prov-DM, these two events don't have temporal information. Thus, in PROV-O, we could represent such information that cannot be expressed in PROV-DM. But so what? I don't think we agreed to the constraint that everything one can express in PROV-O can also be expressed in PROV-DM; the goal of the mapping was just to show how to express (almost) everything in PROV-DM in RDF. > > If you think PROV-O should not be able to express times for association and delegation because PROV-DM cannot, please raise against PROV-O. > > If you think there is a round-tripping property the mapping should have that it doesn't have, please formulate and raise it as a separate issue against the mapping. (This could ultimately imply changes to several things, so the mapping is an appropriate place to raise it.) > > Proposal: Raise question whether Association and Delegation should have time information against PROV-DM; no change needed to mapping. > > --James > Again, it's an interoperability issue, as above. More fundamentally here, I have no idea what this temporal information means. Luc -- Professor Luc Moreau Electronics and Computer Science tel: +44 23 8059 4487 University of Southampton fax: +44 23 8059 2865 Southampton SO17 1BJ email: l.moreau@ecs.soton.ac.uk United Kingdom http://www.ecs.soton.ac.uk/~lavm
Received on Friday, 17 February 2012 15:04:53 UTC