LISTSERV mailing list manager LISTSERV 16.0

Help for BIBFRAME Archives


BIBFRAME Archives

BIBFRAME Archives


BIBFRAME@LISTSERV.LOC.GOV


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

BIBFRAME Home

BIBFRAME Home

BIBFRAME  November 2014

BIBFRAME November 2014

Subject:

Cataloging applications Was: [BIBFRAME] Closed and Open Assumptions was [BIBFRAME] [Topic] Types

From:

"[log in to unmask]" <[log in to unmask]>

Reply-To:

Bibliographic Framework Transition Initiative Forum <[log in to unmask]>

Date:

Sun, 9 Nov 2014 10:35:31 -0500

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (14 lines)

From my point of view, the way to describe the problem here is not "how to help a user fill out a copy cataloging form", because I think we want to avoid most copy cataloging in the Linked Data context.

In the application I would like to imagine, the user might, as you say, input some value (like an ISBN) in order to catalog a resource identified thereby, and instead of having fields auto-filled with information from some other appropriate resource (like a Work), a link would be made from the resource being cataloged to that other resource. The real problem, it seems to me, is how to offer a good selection of linkable resources, and that could be, as you say, resolved using types. It could also be resolved using partitioned local data stores. It could be resolved using external services that do inferencing to discover types, which would allow the local application to remain "dumb" about inferencing while taking advantage of it on behalf of its users. It could be resolved using some other means of identifying "link candidates", like approved lists. I'm not suggesting that selection by type isn't a very obvious choice, but there are others that would be useful in specific situations. It's going to be the case that, for example, a particular workflow for cataloging instances will have restrictions that will make most resources that happen to be works inappropriate suggestions, and other techniques will need to come into play.

None of this argues against the provision of type information. But I don't think it's going to be enough for this kind of utility, and I don't think this kind of utility (important as it is) is a good argument for committing Bibframe to inferencing applications.

---
A. Soroka
The University of Virginia Library

On Nov 7, 2014, at 5:47 PM, Karen Coyle <[log in to unmask]> wrote:

> Where I see a utility in declaring types for resources and having separate resources for, say, Works and Expressions and Manifestations (or the BF two-entity model) is in selecting data for cataloging or display. I can imagine a cataloging module that allows you get get all of the triples with the subject of class Work when you input an ISBN. This would then fill in the Work parts of the copy cataloging form (or Work and Expression using FRBR). In the example above, and in the RDA unconstrained elements, you would need to select individual elements by name or URI. If there are other ways to do this without the use of types, I would be interested to hear them. I believe that Tom Baker has proposed that this can be done with an Application Profile, but at the moment I'm not remembering any details of how that might work.

Top of Message | Previous Page | Permalink

Advanced Options


Options