Warning:
This wiki has been archived and is now read-only.

PIL OWL Ontology Meeting 2012-01-04

From Provenance WG Wiki
Jump to: navigation, search

Meeting Information

prov-wg - Modeling Task Force - OWL group telecon

Agenda

Attendees

  • Tim
  • James
  • Daniel
  • Satya
  • Stian
  • Stephan
  • Khalid

Summary

We discussed five areas where PROV-O is out of date with PROV-DM. Some of these topics can be done immediately, some require issues to be raised against DM. Each topic is a section below and finishes with either a TODO or OUTSTANDING. The last two sections list the topics we have not yet discussed.

Discussions

http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html

prov:follows and prov:precedes properties missing

  • These relate Events. 
  • "Specifically, follows is a partial order between instantaneous events, indicating that an instantaneous event occurs after another. For symmetry, precedes is defined as the inverse of follows. "
  • "generation of the entity denoted by x precedes the end of a and follows the start of a.  " generation-activity-ordering interpretation.
  • don't Events become time:Instants with inXSDDateTime values? They should.. at least the instantaneousEvents.
  • all Events in DM are instantaneous, Daniel. (True, that was voted recently. Sorry)
  • Activity ordering vs. Event ordering?
  • Activities can have lots of events, Events are partially ordered. Activities are ordered based on the ordering of the Events.
  • Satya: wasScheduledBy (but was renamed)
  • Khalid: wasInformedBy and wasStartedBy
  • TODO: need to add a third abstract relation to (follows, precedes) "equal"

Replace Recipe with Plan

  • TODO: just do it :-)

Add Person, Organization and SoftwareAgent as (disjoint) subtypes of Agent. Add equivalence between prov:Person and foaf:Person 

  • satya: why not other types, like hardware agent?
  • satya: software agent versus activities? one could model activities as agents.
  • daniel: DM says activities cannot be entities.
  • DM: This distinction is similar to the distinction between  'continuant' and 'occurrent' in logic.
  • Daniel: Difficult to distinguish between SoftwareAgent not being an activity.
  • Khalid: Activity is the performance of a task.
  • An activity record is not an entity record. Indeed, an entity record represents an entity that exists in full at any point in its characterization interval, persists during this interval, and preserves the characteristics that makes it identifiable.  Alternatively, an activity in something that happens, unfolds or develops through time, but is typically not identifiable by the characteristics it exhibits at any point during its duration.  This distinction is similar to the distinction between  'continuant' and 'occurrent' in logic
  • DM: "5.2.2 Activity Record
  • In PROV-DM, an activity record is a representation of an identifiable activity, which performs a piece of work.
  • An  activity, represented by an activity record, is delimited by its start  and its end events; hence, it occurs over an interval delimited by two  instantaneous events."
  • Tim: Activities are "stuff that happened" that may have involved some agents
  • OUTSTANDING: Satya's concern about activity versus agent.

Note records are missing in prov-o. Also the property "hasAnnotation"

  • Tim: they renamed Annotations to Notes
  • two constructs: bundle a blue thing AND value judgements
  • this is owl:AnnotationProperty AND owl:Annotation (aka reification)
  • 5.2.4 Note Record
  • As provenance records are exchanged between systems, it may be useful  to add extra-information about such records. For instance, a "trust  service" may add value-judgements about the trustworthiness of some of  the assertions made. Likewise, an interactive visualization component  may want to enrich a set of provenance records with information helping  reproduce their visual representation. To help with inter-operability,  PROV-DM introduces a simple annotation mechanism allowing any  identifiable record to be associated with notes.
  • note(ann1,[ex:color="blue", ex:screenX=20, ex:screenY=30])
  • hasAnnotation(g1,n1)
  • Stian: meta provenance: perhaps if one wants to say "the record wasDerivedFrom(a,b) in account c" is derived from "the record wasDerivedFrom(d,e) in account f" - but owl2 annotations can probably do this better than hasAnnotation
  • OUTSTANDING: hasAnnotation can be used for recording meta-provenance? 

wasAssociatedWith is missing in prov-o

  • are we going to live with this naming?
  • did it REPLACE wasControlledBy? So wasAssociatedWith is elaborating wasControlledBy. wasControlledBy is no longer in DM
  • How does one distinguish among the types of responsibilities?
  • Stephan: replaced hadParticipation, and wasAssociatedWith became a superproperty of hadParticipation and wasControlledBy.
  • Tim: so we have to reappy the qualifiedinvolvement reification pattern to model this? Daniel: yes.
  • Stephan: yes
  • Tim: should we propose to reuse wasAssociatedWith's semantics for hadParticipation (i.e., rename wasAssociatedWith to hadParticipation)?
  • nb. most of these wasAssociatedWiths should be property chains and inferred.
  • 5.3 table wasAssociatedWith between activities and agents.
  • Did we already model this with http://www.w3.org/2011/prov/wiki/Qualifed_Involvements_in_PROV-O#Qualified_Participation_with_Roles?
  • Stephan: I think so
  • Daniel: I think so too.
  • TODO: develop proposal to rename wasAssociatedWith to qualified participation.

Did not get to in this telecon

  •  wasStartedBy and wasEndedBy relationships missing in prov-o.  
  •  Note (Daniel G): should wasAssociatedWith, wasStartedBy and  wasEndedBy have qualified involvements? According to prov dm, they can  be qualified with roles. In particular, wasAssociatedWith can be  qualified with a plan: an agent executes an activity according to a  plan. Should the plan be linked to the associatedwith or to the  activity? Can activities have more than 1 plan? (To be discussed). 
  •  actedOnBehalfOf relationship missing. 
  •  prov:steps property (to qualify derivations) is missing. This implies to create a qualified involvement for the derivation too. 
  •  alternateOf and specializationOf properties are missing in prov-o. 
  •  ProvenanceContainer should be renamed RecordContainer. Also,  according to prov-dm a record container should not be defined as a  prov-dm record (i.e, we should not include it in the ontology), because  otherwise it could appear arbitrarily nested inside accounts.  
  •  I think we should remove this term (Daniel G) 
  •  tracedTo relationship missing. 
  •  wasControlledBy, wasEventuallyDerivedFrom, dependedOn,  hadParticipant and wasScheduledAfter properties should be removed from  prov-o. 

Initial list (This is covered in the list above)

  •  Agents association 
  •  Plan (in place of Recipe) 
  •  wasScheduledAfter - overloaded property 
  •  Start/End of Activity 
  •  onBehalfOf relation between agents and an activity 
  •  Derivation - updated in DM, three kinds of derivation