This is a draft for discussion

Purpose of this document

PROV-DM specifies a data model for provenance (realizable using PROV-N, PROV-RDF, or PROV-XML). However, nothing in [PROV-DM part 1] forces such provenance data to be meaningful, that is, to correspond to a consistent history of objects and interactions. Furthermore, provenance records support a number of standard inferences that would be useful to applications or users.

This document specifies _inferences_ over PROV-DM data that applications may employ, including definitions of some provenance expressions in terms of others, and also defines a class of _valid_ PROV-DM data by specifying _constraints_ that valid PROV-DM data must satisfy. Validity is optional, but encouraged, in order to increase the usefulness of provenance.

The specification lists inferences and definitions together in one section (Section ??), and then considers two kinds of validity constraints (Section ??): _structural constraints_ that prescribe , and _event ordering_ constraints that require that the records in a PROV-DM instance are consistent with a sensible ordering of events relating the activities, entities and actors involved. In separate sections we consider additional constraints specific to accounts and collections (Section ?? and ??).

The specification also describes how the inferences, definitions, and constraints should be used (Section ??). Briefly, a PROV-DM compliant application is allowed (but not required) to treat two PROV-DM instances as the same if they are equal after applying the inference rules and possibly reordering expressions, and we can define a canonical form for PROV-DM instances obtained by applying all possible inference rules. In addition, a validating PROV-DM application is required to check that the constraints are satisfied in (the canonical form of) provenance data generated or consumed by the application.

Finally, the specification includes a (non-normative) section (Section ??) describing the rationale for the constraints, particularly background on events, attributes, the the role of inference, and accounts. A formal mathematical model that further justifies the constraints and inferences is found in [PROV-SEM].

Audience

The audience for PROV-DMC is the same as for PROV-DM: developers and users who wish to create, process, share or integrate provenance records on the (Semantic) Web. Not all PROV-compliant applications need to check validity when processing provenance, but many applications could benefit from the inference rules specified here. Conversely, applications that create or transform provenance should try to produce valid provenance, to make it more useful to other applications.

Outline:

1. Introduction/purpose/audience

2. Inferences (cf. old sec. 3)

2.1. Basic inferences

2.2. Definitions

3. Structural constraints (cf. old sec. 3, 6)

4. Event ordering constraints (cf. old sec. 5)

5. Account constraints (cf. old sec. 4)

6. Rationale (cf. old sec. 2).

A. Acknowledgments

B. References