A | B | C | D | E | F | G | H | I | J | |
---|---|---|---|---|---|---|---|---|---|---|

1 | Type | No | Name | Detailed | To be included? | Where? (prov-o.owl, prov-constraints.owl, rules.sparql, n/a) | Who? | How? | Comments? | Jun's proof of concept using Pellet |

2 | Definition | 1 | optional-identifiers | No | Implied by RDF | |||||

3 | Definition | 2 | optional-attributes | No | Implied by RDF | |||||

4 | Definition | 3 | definition-short-forms | See comment | Implied by RDF | Need: wasGeneratedBy <-> qualifiedGeneration etc; Cannot be expressed using OWL (Jun) | ||||

5 | Definition | 4 | optional-placeholders | Yes | prov-constraints.owl | exists/cardinality | Jun: I am not sure we need existential quantification here. Which property do you refer to specifically? | |||

6 | Inference | 5 | communication-generation-use-inference | wasInformedBy (used,wasGeneratedBy) | ? | rule | cannot be implemented in owl using property chain | |||

7 | Inference | 6 | generation-use-communication-inference | ? | rule | No | ||||

8 | Inference | 7 | entity-generation-invalidation-inference | If it's an entity, there is a generation and invalidation | Yes | prov-constraints.owl | exists/cardinality on qualifiedGeneration/Invalidation | Jun: why on these properties, but not on the Entity class? | No | |

9 | Inference | 8 | activity-start-end-inference | If it's an activity, there is a start and end | Yes | prov-constraints.owl | exists/cardinality on qualifiedStart/End | Jun: why not on the Activity Class? | No | |

10 | Inference | 9 | wasStartedBy-inference | wasStartedBy->wasGeneratedBy | ? | property chain or rule | Jun: Does wasStartedBy (a,e, a1) also imply used(a,e) and wasGeneratedBy(e,a1)? Why part of it was not covered in James' rules? More discussion. Khalid: the entity e seems to be optional according to prov-dm, which means teh inference can only be made when e is specified. | No? | ||

11 | Inference | 10 | wasEndedBy-inference | wasEndedBy->wasGeneratedBy | ? | property chain or rule | Same as above | No? | ||

12 | Inference | 11 | derivation-generation-use-inference | wasDerivedFrom->used, wasGeneratedBy | ? | property chain or rule | Jun: suggest as property chain in the owl. Khalid: in the definition of derivation wasDerivedFrom(id; e2, e1, a, g2, u1, attrs), the activity a is option, which means the the inference can only be made when a is specified. | |||

13 | Inference | 12 | revision-is-alternate-inference | wasRevisionOf->alternateOf | ? | rule. Khalid: subPropery? | Khalid: If that inference always holds, why dont we define alternateOf as a subproperty of wasRevisionOf? If so then we should raise this to the group. | |||

14 | Inference | 13 | attribution-inference | wasAttributedTo->wasGeneratedBy, wasAssociatedWith | ? | property chain or rule | Jun: suggest as property chain in the owl. Khalid: the attribution does not specify the activity whereby the entity was generated. So this inference can not be made in OWL. | |||

15 | Inference | 14 | delegation-inference | actedOnBehalfOf->wasAssociatedWith | ? | rule | Jun: not sure how this can be properly implemented. Khalid: This inference can only be made when actedOnBahald specifies the activity, which is an optional attribute. | |||

16 | Inference | 15 | influence-inference | Yes | prov-o.owl | subproperty | Jun: to be raised to the group. Khalid: I agree with defining it as a sub-property. | |||

17 | Inference | 16 | alternate-reflexive | Yes | prov-constraints.owl | reflexive | ||||

18 | Inference | 17 | alternate-transitive | Yes | prov-constraints.owl | transitive | ||||

19 | Inference | 18 | alternate-symmetric | Yes | prov-constraints.owl | symmetric | ||||

20 | Inference | 19 | specialization-transitive | Yes | prov-constraints.owl | transitive | ||||

21 | Inference | 20 | specialization-alternate-inference | Yes | prov-o.owl | subproperty | Jun: to be raised to the group | |||

22 | Inference | 21 | specialization-attributes-inference | ? | rule | Jun: create a class Specialization as a subClass of Entity? Use specializationOf to define an existential quantification? Khalid: not sure this can be done simply by sub-classing. Sometimes the attributes can be heterogeneous, in the sense that they use different datatype/representation. | ||||

23 | Inference | 22 | mention-specialization-inference | Yes | prov-o.owl | subproperty | Jun: to be raised to the group. Khalid: according to the OWL ontology mentionOf is a sub-property of specializationOf | |||

24 | Constraint | 23 | key-object | ? | ||||||

25 | Constraint | 24 | key-properties | ? | ||||||

26 | Constraint | 25 | unique-generation | entity, wasGeneratedBy | Yes | prov-constraints.owl | cardinality quantification on Entity class (max) and rules? | see also inference 8. Khalid: Cant we simply define generation as functional? | ||

27 | Constraint | 26 | unique-invalidation | entity, wasInvalidatedBy | Yes | prov-constraints.owl | cardinality quantification on Entity class (max) and rules? | see also inference 8. Khalid: the same as above | ||

28 | Constraint | 27 | unique-wasStartedBy | activity, wasStartedBy | Yes | prov-constraints.owl | cardinality quantification on Activity class (max) and rules? | see also inference 9. Khalid: the same as above | ||

29 | Constraint | 28 | unique-wasEndedBy | activity, wasEndedBy | Yes | prov-constraints.owl | cardinality quantification on Activity class (max) and rules? | see also inference 9. Khalid: the same as above | ||

30 | Constraint | 29 | unique-startTime | No | rule? | Khalid: define the property as functional? | ||||

31 | Constraint | 30 | unique-endTime | No | rule? | Khalid: define the property as functional? | ||||

32 | Constraint | 31 | unique-mention | ? | I am not sure I agree with this constraint. Didn't review it because it was associated with a feature at risk | |||||

33 | Constraint | 32 | start-precedes-end | No | rule | Khalid: OWL does not seems fit for expressing temporal relationship. so it is better to define them using other mechanisms. | ||||

34 | Constraint | 33 | start-start-ordering | No | rule | |||||

35 | Constraint | 34 | end-end-ordering | No | rule | |||||

36 | Constraint | 35 | usage-within-activity | No | rule | |||||

37 | Constraint | 36 | generation-within-activity | No | rule | |||||

38 | Constraint | 37 | wasInformedBy-ordering | No | rule | |||||

39 | Constraint | 38 | generation-precedes-invalidation | No | rule | |||||

40 | Constraint | 39 | generation-precedes-usage | No | rule | |||||

41 | Constraint | 40 | usage-precedes-invalidation | No | rule | |||||

42 | Constraint | 41 | generation-generation-ordering | No | rule | |||||

43 | Constraint | 42 | invalidation-invalidation-ordering | No | rule | |||||

44 | Constraint | 43 | derivation-usage-generation-ordering | No | rule | |||||

45 | Constraint | 44 | derivation-generation-generation-ordering | No | rule | |||||

46 | Constraint | 45 | wasStartedBy-ordering | No | rule | |||||

47 | Constraint | 46 | wasEndedBy-ordering | No | rule | |||||

48 | Constraint | 47 | specialization-generation-ordering | No | rule | |||||

49 | Constraint | 48 | specialization-invalidation-ordering | No | rule | |||||

50 | Constraint | 49 | wasAssociatedWith-ordering | No | rule | |||||

51 | Constraint | 50 | wasAttributedTo-ordering | No | rule | |||||

52 | Constraint | 51 | actedOnBehalfOf-ordering | No | rule | |||||

53 | Constraint | 52 | typing | Yes | prov-constraints.owl | owl reasoning (possibly) | ||||

54 | Constraint | 53 | impossible-specialization-reflexive | Yes | prov-constraints.owl | irreflexive | ||||

55 | Constraint | 55 | impossible-property-overlap | identifiers of some properties cannot overlap | No | rule. Khalid: in OWL it is possible to define that two classes are disjoint. Is there any similar mechanism that we can use for properties? | ||||

56 | Constraint | 56 | impossible-object-property-overlap | identifiers not overlapping | No | rule | ||||

57 | Constraint | 57 | entity-activity-disjoint | Yes | prov-o.owl? | disjoint classes | Do we implement this in provo.owl? | |||

58 | Constraint | 58 | Membership-empty-collection | Yes | prov-constraints.owl | cardinality quantification (min=0;max=0) | I am not quite sure about this | |||

59 | ||||||||||

60 |

Loading...