W3C home > Mailing lists > Public > public-prov-wg@w3.org > January 2012

RE: PROV-ISSUE-230 (Name-scoping): Name scoping in DM is wrong concept [prov-dm]

From: Myers, Jim <MYERSJ4@rpi.edu>
Date: Tue, 31 Jan 2012 14:39:45 +0000
To: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>, Provenance Working Group WG <public-prov-wg@w3.org>
Message-ID: <3131E7DF4CD2D94287870F5A931EFC2302A56689@EX14MB2.win.rpi.edu>

>>> Are you saying that a PROV account of a particular entity, made in
2012 by asserter X, must be compatible with whatever resolving the
entity's URI in 2020 will tell us about the resource? Or with what the
entity's URI might have resolved to in 2012? As requested by whom,
how?

For the URI of an entity, I think the answer has to be yes - what would it mean to assert that e1 was created by a certain activity if e1 is allowed to become something different in a few years?

The way I look at it, URIs that change over time, or are ambiguous ala www.example.com are identifiers of 'things' and you address the challenges you describe by asserters making new entities that characterize that thing in different ways (or the same way at different times - two entities characterizing the company owning www.example.com asserted at different times don't necessarily refer to the same company.)

The ability to mint new entities that characterize the same thing (e.g. www.example.com) seems to give you all the flexibility needed to address the potential shifts of what the thing means. I expect the major use case will be this one for all the reasons you state with most asserters minting new entities rather than re-using an existing entity (via its global ID),and with provenance users looking for entities characterizing the same thing and making judgments about whether/how to interpret the multiple characterizations of the thing to decide how the information from multiple accounts can be combined in a rational fashion after the fact.

But when asserters do reuse the entity, I think it has to be taken as an assertion that they are truly talking about the same characterization of the same thing, not one that morphs over time. I think we do have that use case as well - multiple people who saw 'Luc-in-Boston' and have tales to tell would want to re-use that specific entity and its URI as an assertion that they were all talking about the same guy on the same trip - if we just had e0 is Luc with location Boston, e1 for luc on business trip X, etc. we can't be sure that e0 and e1 truly represent the same characterization of Luc - reusing e0 in the second account is the way to assert that (when that is what you mean).

  Jim
Received on Tuesday, 31 January 2012 14:40:20 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:58:11 UTC