$5 #361
Replies: 28 comments 49 replies
-
|
In the issue #345, I pasted in my notes from searching through all the tags and identifying which ones have $5. Will have more to say later. Many, but not all, contexts point to an Item-level descriptive element by specific definition. Some may not be so transparent but indicate an institution-specific use of a field. |
Beta Was this translation helpful? Give feedback.
-
|
So this points to two things: |
Beta Was this translation helpful? Give feedback.
-
|
Will probably have to make decisions about ambiguous $5's similar to ones where we map to "lowest level" - which in this case would be item. Thinking about 700 with an institution in $5 and $e former owner etc. |
Beta Was this translation helpful? Give feedback.
-
|
I'm not there yet, but it looks like the whole mapping of this note is conditional on presence of a $5. If there is one, the most likely relation is "Note on item" http://rdaregistry.info/Elements/i/P40028 (coining an IRI for "an item" that's otherwise unspecified), for the $a, and "Current owner collective agent" http://rdaregistry.info/Elements/i/P40117) for the $5. I see that there's a mention of "Holding" in RDA e.g. item property "Holding of" http://www.rdaregistry.info/Elements/i/#P40161 but there doesn't seem to be an entity/ IRI for a "holding". Its definition "A manifestation of a collection work that includes the item." If someone can elucidate whether the meaning of "collection work" could include a library's "holdings" in its collection(s), or whether something else is meant, that could maybe help determine whether this could be useful |
Beta Was this translation helpful? Give feedback.
-
|
Also, The MARC Code List for Organizations https://www.loc.gov/marc/organizations/org-search.php does not provide a URI identifying the organizations. id.loc.gov doesn't provide a web interface or advice on formulating a URI related to the code (e.g. geu for Emory University). It refers to the 40,000+ organizations and downloadable batches but not sure there would be URIs there. |
Beta Was this translation helpful? Give feedback.
-
|
The RDA 'holding of' property relates an item to a collection in which it is held. The collection is represented as a manifestation of a 'collection' category of work that is similar to 'aggregating work', 'diachronic work', and other categories. A collection work has a 111 WEMI structure: I collection work with 1 collection expression that does not require description with 1 collection manifestation and 1 collection item. (Do not confuse collection item with collected item aka holding.). |
Beta Was this translation helpful? Give feedback.
-
|
Also, see CONSER Editing Guide for how $5 is used in field 588: https://www.loc.gov/aba/pcc/conser/CEG/588.doc
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Wednesday, June 1, 2022 7:14 AM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Subscribed ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
Tag contexts where the entity related to $5 is almost certainly an Item or Items:
* 026 Fingerprint identifier
* 037 - Source of Acquisition
* 506 - Restrictions on Access Note
* 533 - Reproduction Note
* 541 - Immediate Source of Acquisition Note
* 561 - Ownership and Custodial History
* 562 - Copy and Version Identification Note
* 563 - Binding Information
* 583 - Action Note
* 584 - Accumulation and Frequency of Use Note (note: rate of accumulation might be a manifestation-level aspect for continuing resource)
* 585 - Exhibitions Note
* 843 - (Holdings) Reproduction Note
* 845 - (Holdings) Terms Governing Use and Reproduction Note
Tag contexts where the entity related to $5 is not clear (at least to me). These are likely to relate to an item (which is why the $5 provides localization) but may relate to something else (manifestation, metadata provenance, etc.):
* 246 - Varying Form of Title
* 500 - General Note
* 501 - With Note
* 538 - System Details Note (probably related to manifestation, but could be unique to item?)
* 540 - Terms Governing Use and Reproduction Note (may apply to manifestation or item?)
* 588 - Source of Description Note (relates to the metadata)
* 655 - Index Term-Genre/Form
* 700 - Added Entry-Personal Name
* 710 - Added Entry-Corporate Name
* 711 - Added Entry-Meeting Name
* 730 - Added Entry-Uniform Title (R) Uniform title
* 740 - Added Entry-Uncontrolled Related/Analytical Title
* 758 - Resource Identifier
* 800 - Series Added Entry-Personal Name
* 810 - Series Added Entry-Corporate Name
* 811 - Series Added Entry-Meeting Name
* 830 - Series Added Entry-Uniform Title
* 880 - Alternate Graphic Representation
* 885 - Matching Information (related to cataloging)
* 886 - Foreign MARC Information Field (related to metadata conversion)
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVB46MTATUR2U3LDVW43VM5V3RANCNFSM5WQBZS4Q>.
You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
In this case, it's clear that the $5 relates to provenance - of the note giving information based on special knowledge or judgment (which is probably a note on manifestation...?). |
Beta Was this translation helpful? Give feedback.
-
|
Actually, see: https://id.loc.gov/vocabulary/organizations.html
Code List for Cultural Heritage Organizations - LC Linked Data Service: Authorities and Vocabularies - Library of Congress<https://id.loc.gov/vocabulary/organizations.html>
The Linked Data Service provides access to commonly found standards and vocabularies promulgated by the Library of Congress. This includes data values and the controlled vocabularies that house them. Datasets available include LCSH, BIBFRAME, LC Name Authorities, LC Classification, MARC codes, PREMIS vocabularies, ISO language codes, and more.
id.loc.gov
URI for Emory University: http://id.loc.gov/vocabulary/organizations/geu
Emory University - LC Linked Data Service: Authorities and Vocabularies | Library of Congress<http://id.loc.gov/vocabulary/organizations/geu>
The Linked Data Service provides access to commonly found standards and vocabularies promulgated by the Library of Congress. This includes data values and the controlled vocabularies that house them. Datasets available include LCSH, BIBFRAME, LC Name Authorities, LC Classification, MARC codes, PREMIS vocabularies, ISO language codes, and more.
id.loc.gov
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Wednesday, July 13, 2022 7:44 AM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Comment ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
Also, The MARC Code List for Organizations https://www.loc.gov/marc/organizations/org-search.php does not provide a URI identifying the organizations. id.loc.gov<http://id.loc.gov/> doesn't provide a web interface or advice on formulating a URI related to the code (e.g. geu for Emory University). It refers to the 40,000+ organizations and downloadable batches but not sure there would be URIs there.
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVB26BYKFHYINEFTF73TVT3I4FANCNFSM5WQBZS4Q>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Thanks, I had found it - it just wasn't where I was looking. Very good, we can use the URIs. |
Beta Was this translation helpful? Give feedback.
-
|
Well, except they are SKOS/MADS URIs for the authority files with no RWO that I can see. :-( |
Beta Was this translation helpful? Give feedback.
-
|
The entity represented by the LoC URI, for example for the cultural heritage organization code NjP, is a union of madsrdf:CorporateName, madsrdf:Authority, and skos:Concept. This is not an RDA Corporate Body, and cannot be the object of an RDA element with range Agent/Corporate Body. madsrdf:Authority is declared a subclass of skos:Concept, so this can be simplified to the union of madsrdf:CorporateName and skos:Concept. madsrdf:CorporateName is a subclass of madsrdf:Name is a subclass of madsrdf:SimpleType ("a resource with a label constituting a single word or phrase") is a subclass of madsrdf:MADSType (top of the hierarchy, but no definition). The best fit with RDA seems to be Nomen: madsrdf:Name ("a resource whose label represents a name") is equivalent to Nomen ("a label for any RDA entity except a nomen"), and both are similar to skosxl:Label (a lexical entity with a single literal form). |
Beta Was this translation helpful? Give feedback.
-
|
A general pattern for dealing with $5 is: ex:Manifestation1 rdamo:P30103 ex:Item1 . ex:Manifestation1 is the manifestation described by the record. ex:CollectionManifestation/{$5} is the manifestation of a collection work; the IRI has the institution code as its local part. The appellation of the manifestation is constructed from the label corresponding to the insititution code. It is not a title because it is supplied by the cataloguer. It could be refined as an access point, using an inverted string encoding scheme of {$5 label} + ". Collection", but I'm not sure that the labels are authority-controlled. "Library collection of ..." might be better boilerplate in general, but could lead to slightly weird or redundant results such as "Library collection of X Library." ex:Item1 is an item that exemplifies the manifestation. The following worked examples are taken from the MARC 21 Bibliographic manual. Example 1: 500 ##$aPlates 4, 5, and 9 are wanting.$5DLC ex:Manifestation1 rdamo:P30103 ex:Item1 . Note that rdaid:P40003 "has modification of item" is a more accurate property to use, but it is not possible to determine if the statement is about a modification. This is a subproperty of rdaid:P40028. Example 2: 500 ##$aUniversity Library's copy has ms. notes by author on endpapers.$5WU ex:Manifestation1 rdamo:P30103 ex:Item1 . This is also a modification, but again it cannot be determined by simple machine parsing. Example 3: 710 2#$aBridgewater Library,$dformer owner.$5NjP ex:Manifestation1 rdamo:P30103 ex:Item1 . The note includes other subfields using standard boilerplate and punctuation. $d is defined as date of meeting, etc., so I assume this is an error, and it should be $e (relator term). $a on its own results in a note that is too vague and potentially misleading. Example 4: 583 ##$aQueued for preservation$c19861010-$ePriority$fTitle IIC project$5DLC ex:Manifestation1 rdamo:P30103 ex:Item1 . The note is a more elaborate example of including multiple subfields using a string encoding scheme. The last statement has already been made for example 1. It is feasible to generate this statement in advance for all LC codes, rather than generate it for every instance of the tag. Tag 583 is unambigously focussed on an item, so a more refined model can be developed for the tag, albeit as an extension of RDA. The 3R Project discussed this with the rare books community, based on a model for 'transaction history' of an item. This includes transactions such as acquistion and transfer of ownership, nature of acquisition (donation, purchase, permanent loan, temporary loan, etc.), and collecting history or bibliographic provenance (as in example 3), plus date of transaction and agents involved in the transaction. This could replace the broad element rdai:P40026 "has custodial history of item", but the community advised keeping it simple for the time being (this was 5 years ago). I'm wondering if this would be worth raising this again in the context of our work, the newly-published official RDA extension for collection description, and DCRMRDA. This may not be immediately relevant to our work as it will probably take 18 months minimum from asking the RDA Technical Committee to draft proposals and having them approved and added to RDA. |
Beta Was this translation helpful? Give feedback.
-
|
Could this example ex:Manifestation1 rdamo:P30103 ex:Item1 . be rewritten to retain the Corporate Body entity and its relation to the item separrately, so the URI for the corporate body could be used? I assume so but haven't time to tackle. |
Beta Was this translation helpful? Give feedback.
-
|
important note from meeting: certain relationship designators can be unambiguously mapped to an item entity |
Beta Was this translation helpful? Give feedback.
-
|
I think these are the MARC 21 relators that indicate an item2agent relationship: Binder [bnd] Conservator [con] Depositor [dpt] Donor [dnr] Former owner [fmo] Lender [len] Owner [own] Restorationist [rsr] Seller [sll] The relators fall into two distinct categories: Modifier of an item: Owner of an item: |
Beta Was this translation helpful? Give feedback.
-
|
@lake44me asks if the 'Bridgewater Library' example can be rewritten to retain the 'Corporate Body' entity - but as discussed above, the LoC URI represents a nomen, not a corporate body. To link to the LoC data, we still have to mint a URI for the corporate body: ex:CorporateBodyAgent/NjP rdaao:P50380 http://id.loc.gov/vocabulary/organizations/njp . I think it is only safe to say that an item that is held in a collection manifestation is 'collected' by an agent; in general, it is not safe to say that it is 'owned' by an agent because there is no intrinsic relationship between collecting and ownership (e.g. deposit and lending activities; exhibitions = items on loan from multiple agents; etc.). Of course, the existence of an 'owner' relator removes the uncertainty. In this example, the 'former owner' relator is an ownership relationhip, but we have to make an additional assumption that the organization that makes the statement is the current owner; that is, Princeton University (Library) is the current owner (the semantics of the LoC relators is a bit fuzzy). Of course, the item could now have a new 'current' owner, but all previous owners were/are current owners at some point in time and it is not false to say so. If this assumption is safe, we can say: ex:Item1 rdaio:P40021 ex:CorporateBodyAgent/NjP . This directly relates the item to its owner. As discussed on the call, these statements can be added to the example without replacing any statements and without resulting in an incoherent description set. A more general extension to the example description set avoids ownership assumptions. The RDA 'collector' relationship is between a collection work and an agent. We need to mint a URI for the collection work and relate it to the collection manifestation that holds the item: ex:CollectionManifestation/NjP rdamo:P30135 ex:CollectionWork/NjP . Then relate the collection work to the collector organization (corporate body): ex:CollectionWork/NjP rdawo:P10631 ex:CorporateBodyAgent/NjP . Finally, the corporate body is linked to the LoC nomen-like data: ex:CorporateBodyAgent/NjP rdaao:P50380 http://id.loc.gov/vocabulary/organizations/njp . Again, these statements can be added to the description set without causing incoherency. |
Beta Was this translation helpful? Give feedback.
-
|
This is the basic model of a collection in RDA. Properties without labels are primary/resource properties. The cardinality of each property is indicated. * means 'at any point in time'; over time, the holding collection, collector agent, and owner agent can be replaced by another. The classes with italic labels are not normally identified or described. For 'bulk ownership', an application may choose from the following:
Note that there can only be one owner agent at a time. This reflects the legal context of the relationship: two persons or collective agents who jointly 'own' an item are treated as a single collective agent. Note that there is no direct correlation between 'owner' and 'collector'. The institutional collection of a large library may hold items that are not owned by the institution/library (e.g. on long-term deposit; on short-term loan for a specific event; etc.). This balance is different for archival collections and museum collections. |
Beta Was this translation helpful? Give feedback.
-
|
In attempting to apply Gordon's mapping examples to the tag 500 mapping, I'm feeling uneasy about not actually mapping the code from the Code List for Cultural Heritage Organizations anywhere (well, we could have a convention when minting a URI for a "collection manifestation", but that's not really RDA data. Note that the list is of organizations, not collections. What Gordon suggests (capturing the spelled out name of the organization from the code list and making it part of a "note on manifestation" about the minted collection manifestation, is roundabout and loses the source of the data. It does save us from minting more URIs however. Am I reading RDA correctly that, if we wanted to surface that code as an identifier, it would need to be a Nomen/Nomen string? Hoping there's a better way to finagle this... maybe adding the code to the note? Laura |
Beta Was this translation helpful? Give feedback.
-
|
Last week there was a call for examples of subfield 5 usage. With the help of Lisa Hamlett, our Alma guru, we were able to wrassle out of Alma a file that captures just the record IDs and the fields that contain a subfield 5, which I am attaching. If you want fuller information/record views on any of these, let me know, but it does kind of give a picture of how one institution uses the subfields and where they appear in our records. Any full record can be viewed in Library Search, our Blacklight catalog, by replacing the record MMSID after the slash in this URL https://search.libraries.emory.edu/catalog/990011417900302486 with the 001 number from the report. The Staff View shows the MARC in a simple text format. This illustrates that there's a "local practice" element in how these are used and it varies over time and from library to library... there are also a few that were not added by our catalogers, e.g an online journal available through Directory of Open Access Journals using an Alma "Community Zone" record. 000||00563nam a2200085 a 4500 One observation from just scanning through these - at least in our practices, it looks like a $5 in a 655 Genre/Form field is less likely to indicate a genre or form aspect of a specific copy, than it is to indicate an application of a locally derived vocabulary or local usage of a vocabulary for genre/form. That's just a guess. Sometimes I have a hard time figuring out why the $5's are there. If we could get a rare materials cataloger in on this it could shed some light. Is there a general best practice about placement of $5? I'll tray to catch our rare book cataloger who is leaving at end of next week. There were some $5's that appear on a 590 (local field), and perhaps other places where they aren't valid; things happen and we wouldn't do a mapping for oddities like that anyway, right? |
Beta Was this translation helpful? Give feedback.
-
|
I got this reply today: The 655 contains lcgft, but also RBMS and AAT vocabularies, and the local terms we use to show AfAm creators/publishers. All of those, especially RBMS are almost by definition local copies as they label binding, paper, and provenance (the most common ones we use). RBMS and local terms are generally the ones with $5. I hope that helps. Let me know if you have more questions, Beth Shoemaker |
Beta Was this translation helpful? Give feedback.
-
|
Wanted to add to this discussion, the issue that @CECSpecialistI brought up at the very end of last week's meeting. She mentioned, maybe a note of some kind that actually discussed multiple items held in a collection. Someone else mentioned having multiple tags within a record with a $5 for the same institution. We needn't necessarily assume that the content of these tags all pertains to the same item. Practically speaking, I don't see any way that a conversion program can do anything other than mint one item per $5 code/institution/collection, and assume that any $5's in the record that pertain to item-level description (important distinction - some may have a different meaning) are related to that item. It would serve the role of a generic "there exists at least one item in a collection" item - in other words, a "holding"... and associating it with a collection is a good thing. If it turns out later that multiple items are involved, that would need to be sorted out by hand. Does anyone see it differently? It is theoretically possible for subfield $5 's with different codes to exist in the same record. Never saw this in the wild. Imagining if a particular item changed ownership, the new owner might, for example, note something about the condition of the item after "the unfortunate fire at X library" which had also added a note with a $5... |
Beta Was this translation helpful? Give feedback.
-
|
The record is for the manifestation though, so multiple exemplars of the same manifestation at different institutions have different $5 values, which I have seen in the wild in special collections cataloging. I'm not sure what you're getting at when you say
Are you saying we ought to mint one item per institution indicated by $5 per MARC record, and let that serve as a hybrid item/holding record until individual institutions reconcile their item(s) to the items in the records? Or, are you suggesting that we create one holdings entity, distinct from an item entity, for each institution that has a $5 associated with the MARC record and leave items to individual institutions? |
Beta Was this translation helpful? Give feedback.
-
|
P40161 | "is holding of" is a property that "Relates an item to a manifestation of a collection work that includes the item." That's the only RDA vocabulary term using the word "holding" that I can find. There's no "Holding" entity. We could ask for one or invent it as an extension for this project. Given that limitation, I suggest the best we can do is mint one item per institution code per MARC record, and have it serve for mapping fields where $5 indicates local copy information. Any such mapped field should be reviewed after conversion if possible, and sorted out into multiple item descriptions if needed. It's possible for a MARC record to contain MARC holdings tags (one or more) with detailed item data (https://www.loc.gov/marc/holdings/hd876878.html), although not often seen (it's something that may be included with publication of MARC records for external applications or consumers). But even then, there'd be no way to marry information in a 500, 655 or 710, say, with a $5, to an item description it belongs with, using a program. In the example I used from our report of subfield 5 instances in Emory MARC, when I looked it up, there are multiple copies - some of which are apparently part of the general collection in storage. [i:P40161](http://rdaregistry.info/Elements/i/P40161) "is holding of" |
Beta Was this translation helpful? Give feedback.
-
This seems to be the most prudent course of action to me. |
Beta Was this translation helpful? Give feedback.
-
|
I disagree with 'mint one item per institution' because it results in false statements such as an item being simultaneously acquired/sold. It is better to err on the side of caution and mint an item for every occurrence. Although this results in multiple IRIs for a single item, this can be resolved in some cases with automatic inferencing using additional data, and in all cases with human intervention - and all of the statements are true. We should not be averse to mint-bloat; the recent announcement of DBpedia Day says "... we developed a way to scale up our identifier base into the billions ..." (In my opinion, woefully short of what will be required, but indicative of the trend ;-) We will have to take the same approach when minting IRIs for works (!) and expressions (!!). Mint-bloat is inevitable, and is ultimately resolved through human intervention, aka crowd-sourcing, something which is difficult to achieve via pay-walls and closed-worlds. |
Beta Was this translation helpful? Give feedback.
-
|
Wouldn't we want to get the URI associated with the code?
[cid:783ca204-4b4c-41d7-b140-34163bef54f1]
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Monday, August 8, 2022 2:36 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Comment ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
In attempting to apply Gordon's mapping examples to the tag 500 mapping, I'm feeling uneasy about not actually mapping the code from the Code List for Cultural Heritage Organizations<https://id.loc.gov/vocabulary/organizations.html> anywhere (well, we could have a convention when minting a URI for a "collection manifestation", but that's not really RDA data.
Note that the list is of organizations, not collections.
What Gordon suggests (capturing the spelled out name of the organization from the code list and making it part of a "note on manifestation" about the minted collection manifestation, is roundabout and loses the source of the data. It does save us from minting more URIs however.
Am I reading RDA correctly that, if we wanted to surface that code as an identifier, it would need to be a Nomen/Nomen string?
And if so, would it be more appropriate to make this an identifier of the "Collector Agent" associated with the "Collection Work", both of which would need to be minted?
Hoping there's a better way to finagle this... maybe adding the code to the note?
Laura
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVB43RVNUD4SUNTSDA7LVYF4VDANCNFSM5WQBZS4Q>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Mocking up a possibility: 500 a| Pitts Theology Library copy from the collection of the Christoph Keller, Jr. Library, The General Theological Seminary. 5| GEU-T ex:Manifestation1 rdamo:P30103 ex:Item1 // Manifestation has exemplar of manifestation - a single item - is this the general property for associating an item? ex: Item1 rdaid:P40028 "Pitts Theology Library copy from the collection of the Christoph Keller, Jr. Library, The General Theological Seminary" // item has note on item -> content of subfield a string ex: Item1 rdaio:P40161 ex:CollectionManifestation/GEU-T // Item - is holding of - the Registry definition says "collection work" but Gordon uses collection manifestation. Note that adding the institution code to the minted URI is just a suggested convention ex:CollectionManifestation/GEU-T rdamd:P30277 "Collection of Candler School of Theology (geut in https://id.loc.gov/vocabulary/organizations.html)." //collection manifestation - has appellation of manifestation - "Collection of Candler School of Theology (geut in https://id.loc.gov/vocabulary/organizations.html)." -note: I added the subfield 5 code and its source in parentheses to the note as an option, but observe that the code used in our record's not the same as the code appearing in the id.loc.gov organizations list, nor is that code related to the library specifically - perhaps another list needs to be cited?)** |
Beta Was this translation helpful? Give feedback.
-
|
Maybe we need to map these codes to the RWO URI's from the LCNAF |
Beta Was this translation helpful? Give feedback.
-
|
id.loc.gov data from LCNAF doesn't have RWO URIs does it? Find me one :-). I can track down the PCC Task Group on URIs in MARC report - I think it recommends using subfield 0 (for an authority or system identifier) rather than 1 (for RWO) when using LCNAF URIs. But, I'm in another presentation now so should pay attention. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry I was flat wrong - I forgot about the different URL path for Agent which provides a URI labeled "RWO" for agents. It's LCSH subjects that don't have RWO URIs. Not sure about Author/title authorities but that's being sorted differently, apparently, so maybe RWO URIs for Work (RDA) and Expression (RDA) will emerge in whatever shape those records are transformed into. |
Beta Was this translation helpful? Give feedback.
-
|
The GEU-T form of the identifier is given in the MARC Code List for Organizations. If we reference that, it would be https://www.loc.gov/marc/organizations/org-search.php. GEU-T is listed as Emory University, Candler School of Theology with Variant name(s): Candler School of Theology. That list also gives a normalized form (geut) and an international form id.loc.gov's Cultural Heritage Organizations list has much of the same information. Interestingly, you can't search by the GEU-T form, only the "normalized" geut. Haven't seen mention of an API to query the MARC code list for orgs. There's a tiny bit on the search/query page for id.loc.gov. Basically no query across the internet but you can download the dataset and use various tools on it. It's only about 40,000 records. In last week's meeting, I think it looked like Adam jumped from the id.loc.gov cultural organization page to a NACO Name Authority for the same organization. It went by to fast for me to see the details. Is there a way, are the two "linked", or does another query need to be done based on the appellation/nomen in the organization data? If so, there might be danger trying to introduce the organization NACO ID, and there might be a lack of synch between code list and NACO. The org name or code is not the collection work, etc. Which is why we have to coin a the name for the collection as well as its URI. |
Beta Was this translation helpful? Give feedback.
-
|
As we looked at today, yes the id.loc.gov agent "records" give both a URI for the authority and a RWO URI. For example:
Pitts Theology Library
*
URI(s)
* http://id.loc.gov/authorities/names/n96022791
Additional Information
* http://id.loc.gov/rwo/agents/n96022791
The former is the authority URI (which we would put in $0) and the latter is the RWO URI (which would be put in $1).
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Wednesday, August 10, 2022 8:11 AM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Comment ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
id.loc.gov data from LCNAF doesn't have RWO URIs does it? Find me one :-). I can track down the PCC Task Group on URIs in MARC report - I think it recommends using subfield 0 (for an authority or system identifier) rather than 1 (for RWO) when using LCNAF URIs. But, I'm in another presentation now so should pay attention.
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVBZIWIPSFGNN3FMIYOLVYPBBRANCNFSM5WQBZS4Q>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Cool URIs says "On the Semantic Web, URIs identify not just Web documents, but also real-world objects like people and cars, and even abstract ideas and non-existing things like a mythical unicorn. We call these real-world objects or things." This is a circular definition, and is the cause of much confusion. A Web document is, of course, a RWO itself; in RDA, it is treated as an instance of Item/Manifestation. It then follows that Work and Expression are also RWOs (abstract ideas), as well as names and labels (Nomen). A Web document (RWO) that describes another RWO is either a manifestation of a descriptive work or a metadata work. The difference is that a metadata work is composed of metadata statements that are isomorphic to the triple form subject-predicate-object, and a descriptive work is any other form of description. http://id.loc.gov/authorities/names/n96022791 represents an instance of "A concept with a controlled label" + "Describes a resource whose label is the name of a corporate entity, which may include political or ecclesiastical entities" (the definitions of madsrdf:Authority and madsrdf:CorporateName).
http://id.loc.gov/rwo/agents/n96022791 represents an instance "A madsrdf:RWO is an abstract entity and identifies a Real World Object (RWO) identified by the label of a madsrdf:Authority or madsrdf:DeprecatedAuthority" + "Corporation or group of persons and/or organizations that acts, or may act, as a unit" + ? (the definitions of madsrdf:RWO and bf:Organization; FOAF is not responding). I think the best :-( interpretation is that this is an instance of (RDA) Corporate Body. Back to the first IRI ... SKOS Reference says "A SKOS concept can be viewed as an idea or notion; a unit of thought. However, what constitutes a unit of thought is subjective, and this definition is meant to be suggestive, rather than restrictive. The notion of a SKOS concept is useful when describing the conceptual or intellectual structure of a knowledge organization system, and when referring to specific ideas or meanings established within a KOS. Note that, because SKOS is designed to be a vehicle for representing semi-formal KOS, such as thesauri and classification schemes, a certain amount of flexibility has been built in to the formal definition of this class". This flexibility allows application to 'authority files', but these are not usually treated as KOSs. If the 'resource' is not a corporate entity/Corporate Body, then
I don't think this is intended, because the same treatment is given to Person, and as far as I know, http://id.loc.gov/authorities/names/nb2001072552 is not a concept :-)
If we substitute 'label' for 'concept' and 'resource' in the definitions, we get more circularity and weirdness: "A label with a controlled label" + "Describes a label whose label is the name ...". But I think this is the 'best' interpretation for mapping purposes: http://id.loc.gov/authorities/names/n96022791 isA rdac:C10012 . [I can't see evidence that the IRI represents a 'nomen cluster', which is the nearest LRM/RDA concept to 'authority record'.] |
Beta Was this translation helpful? Give feedback.
-
|
Agreed.
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Crystal Clements ***@***.***>
Sent: Wednesday, August 10, 2022 2:01 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Comment ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
I suggest the best we can do is mint one item per institution code per MARC record, and have it serve for mapping fields where $5 indicates local copy information. Any such mapped field should be reviewed after conversion if possible, and sorted out into multiple item descriptions if needed.
This seems to be the most prudent course of action to me.
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVBZFVK6GJSK32M3OC43VYQKEFANCNFSM5WQBZS4Q>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Laura,
The NACO agents in id.loc.gov are not linked to the Cultural Heritage Organizations. But if you start at the home page of id.loc.gov and just do a keyword search on "Candler School" you get a results page that links to both of them:
[cid:02360070-f650-4a5b-9903-5694628a3b8a]
It also links to the Provider page as well, which has its own identifier and URI. The providers are not linked to authorities or cultural heritage organizations either. The providers were created by extracting publishers from publisher statements in 260/264 $b's. Which means you get URIs for corporate entities such as:
The school committee<https://id.loc.gov/entities/providers/a7d83caa5bf72b29545bc9ccbeeffe14>
The Central high school<https://id.loc.gov/entities/providers/f4d9210467aa06080f505fa49bb62f8d>
Governors of the School<https://id.loc.gov/entities/providers/19f5bb08dcd39877011e95621b7ce198>
Of the Society<https://id.loc.gov/entities/providers/04f2f4ce03f0426ce876215e34d1a13b>
Issued by the Society<https://id.loc.gov/entities/providers/1688b1fcce5ae40cdb7be4992cee171e>
By the Society. Kansas City, Mo<https://id.loc.gov/entities/providers/06215c1730a709447d1dcc94500c6ae4>
And other similar gibberish.
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Monday, August 15, 2022 2:26 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Comment ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
The GEU-T form of the identifier is given in the MARC Code List for Organizations<https://www.loc.gov/marc/organizations/org-search.php>. If we reference that, it would be https://www.loc.gov/marc/organizations/org-search.php. GEU-T is listed as Emory University, Candler School of Theology with Variant name(s): Candler School of Theology. That list also gives a normalized form (geut) and an international form
id.loc.gov's Cultural Heritage Organizations<https://id.loc.gov/search/?q=cs:http://id.loc.gov/vocabulary/organizations> list has much of the same information. Interestingly, you can't search by the GEU-T form, only the "normalized" geut.
Haven't seen mention of an API to query the MARC code list for orgs. There's a tiny bit on the search/query page<https://id.loc.gov/techcenter/searching.html> for id.loc.gov. Basically no query across the internet but you can download the dataset and use various tools on it. It's only about 40,000 records.
In last week's meeting, I think it looked like Adam jumped from the id.loc.gov cultural organization page to a NACO Name Authority for the same organization. It went by to fast for me to see the details. Is there a way, are the two "linked", or does another query need to be done based on the appellation/nomen in the organization data? If so, there might be danger trying to introduce the organization NACO ID, and there might be a lack of synch between code list and NACO. The org name or code is not the collection work, etc. Which is why we have to coin a the name for the collection as well as its URI.
—
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ADFBVB5QYWCYBKWEH5VXEDLVZKYXRANCNFSM5WQBZS4Q>.
You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
I see no one else has taken the bait. Sorry again this is last minute - I am not sure about it but here's a more detailed relationship mapping. ex:Manifestation1 rdamo:P30103 ex:Item1 // Manifestation (minted IRI) has exemplar of manifestation - single item (minted IRI) ex: Item1 rdaid:P40028 "Pitts Theology Library copy from the collection of the Christoph Keller, Jr. Library, The General Theological Seminary" // minted item IRI - has note on item – string, content of subfield a **ex:CollectionManifestation/GEU-T rdamo:P30460 ex:Item1"" // minted collection manifestation IRI – holding – minted item IRI ex:CollectionManifestation/GEU-T rdamo:P30265 ex:CollectionItem/GEU-T //minted collection manifestation IRI – has related item of manifestation – minted collection item IRI ex:CollectionItem/GEU-T rdai:P40021 ex:CollectionOwnerAgent/GEU-T //minted collection item IRI – has owner agent – minted collection owner agent IRI ex:CollectionOwnerAgent/GEU-T rdaa:P50385 ex:Nomen1 //Minted collection owner agent – has name of agent – minted nomen IRI ex:Nomen1 rdand:P80033 “GEU-T” //Minted nomen for collection owner agent – is identifier for agent of – code (content of subfield 5) exNomen1 rdand:P80069 "https://www.loc.gov/marc/organizations/" Minted nomen for collection owner agent – schema of nomen – URL for MARC code list for organizations as a string |
Beta Was this translation helpful? Give feedback.
-
|
Here's a summary of my understanding of where we landed today (2022-08-17). Please correct/give feedback/etc. sometime before next meeting so we can get this recorded :) @AdamSchiff @GordonDunsire @JianPLee @mcm104 @SitaKB @szapoun @gerontakos @pan-zhuo @junghaelee $5 Related to ItemsThis mapping will apply when $5 indicates an item-level statement (most times) Preliminary processing for cultural heritage organizations and their collectionsPrior to transformation:
MARC2RDA MappingDuring transformation, when $5 is encountered and this mapping is indicated,
Illustration of model:
|
Beta Was this translation helpful? Give feedback.
-
|
I have taken a closer look at the LC files relevant to processing $5. The files are:
Test $5 organization code: "NjP" Cultural heritage organizations: uri pattern: http://id.loc.gov/vocabulary/organizations/ + normalized code uri: http://id.loc.gov/vocabulary/organizations/njp skos:notation "NjP"^^http://id.loc.gov/datatypes/orgs/code . // $5 value skos:prefLabel "Princeton University" // no 'library' qualifier Name authority file: uri pattern: http://id.loc.gov/authorities/names/ + authority number Search: princeton university library uri: http://id.loc.gov/authorities/names/n82003163 skos:prefLabel "Princeton University. Library" . skos:altLabel "Princeton University. Harvey S. Firestone Memorial Library" . madsrdf:identifiesRWO http://id.loc.gov/rwo/agents/n82003163 . // "Associates a madsrdf:Authority with the Real World Object that is the subject of the authority's label" RWO/Agents (id.loc.gov general search) uri pattern: http://id.loc.gov/rwo/agents/ + authority number uri: http://id.loc.gov/rwo/agents/n82003163 rdfs:label "Princeton University. Library" . madsrdf:entityDescriptor http://id.loc.gov/authorities/subjects/sh85076584 . madsrdf:isIdentifiedByAuthority http://id.loc.gov/authorities/names/n82003163 . // "Associates a Real World Object with its Authority description" Discussion: madsrdf semantics indicate that the equivalent RDA mappings are: madsrdf:identifiesRWO = rdano:P80144 // "is preferred name of corporate body of" http://id.loc.gov/authorities/names/n82003163 rdano:P80054 http://id.loc.gov/rwo/agents/n82003163 Mapping 1: To include the preferred label from the Agent data, treat it as a nomen string: Map skos:prefLabel to rdand:P80068 "has nomen string" http://id.loc.gov/authorities/names/n82003163 skos:prefLabel "Princeton University. Library" . Mapping 2: To include the alternate labels from the Agent data, treated them as variant names of the corporate body:
http://id.loc.gov/authorities/names/n82003163 skos:altLabel "Firestone Library" . Mapping 3: To include the Agent subject, treat it as a category of corporate body: Map madsrdf:entityDescriptor to rdaao:P50237 "has category of corporate body". It is best to use the object property with the subject authority uri instead of de-referencing or mapping the heading because the uri indicates the categorization system and can be de-referenced for more than just the category term. http://id.loc.gov/rwo/agents/n82003163 madsrdf:entityDescriptor http://id.loc.gov/authorities/subjects/sh85076584 . Issue 1: There is no extrinsic link between the Cultural heritage organizations and the Name authorities. Searching the Name authorities for "Princeton University" (the unprocessed label) gives 500+ hits. Seach for "Princeton University Library" (adding a key word to the label) gives 49 hits. Using a more precise example (narrower), searching for "Princeton University, Gest Library" gives 4 hits, the nearest of which is the name of the collection, not the organization ... It will be possible to generate a default collection manifestation and work from the Cultural heritage organization data, and to generate RDA-conformant corporate bodies and nomens from the Name authority + RWO/Agent data, but the collection work cannot be linked to the corporate body owner without human intervention. One possible way of doing this is to process both ends of the chain and publish the two graphs in a crowdsourcing application (wikidata?). Note that the corporate body/name graph is a considerably bigger super-set of cultural organizations/collections. Download sizes are 9 MB for the skos n-triples version of Cultural heritage organizations (the mads data is not relevant), and 7 GB for the skos/mads version of Name authorities. Issue 2: The RWO/Agents file has no open downloads. If the data cannot be downloaded, it cannot be processed offline. Instead, online de-referencing and processing of the Agent uri is required. Issue 3: The semantics of the LC instances are unclear, and it is difficult and unsafe to assume that they directly correspond to RDA/LRM entities. It is safer to replace the LC uris with local transform uris. The transform can use the LC local part (holdings code or authority number) added to the transform base domain, and the pattern mappings can be published as an alignment for future application. |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI and others - Here's my draft attempt at a worked-out example of RDA for multiple $5 subfielded fields in an example record. I got a bit shaky when I got to the nomen part at the end 'cause I have not dealt with them before. Please review carefully and suggest changes - to the mapping, the notes, or the formatting of examples. `Fragments of Emory held MARC record containing multiple $5: 001 990000266450302486 RDA statements: ex:Manifestation1 rdamo:P30103 ex:Item1* ex:Item1 rdaid:P40028 "Initial letter E illuminated in blue, green, gold, purple and black; elaborately decorated at foot of page with coat of arms. Initials supplied in red." ex:Manifestation1 rdamo:P30103 ex:Item2 ex:Item2 rdaid:P40028 "Frequent annotations, no longer legible because of washing." ex:Manifestation1 rdamo:P30103 ex:Item3 ex:Item3 rdaid:P40028 "From the collection of Charles Spenser, 3rd Earl of Sunderland (Puttick and Simpson, Sale II, April 22, 1882, lot 4351)." ex:Manifestation1 rdamo:P30103 ex:Item4 ex:Item4 rdaio:P40019 ex:Agent1 ex:Agent1 rdaao:P50403 ex:Nomen1 // or id.loc.gov agent IRI **for each of the minted items, generate more statements following this pattern: ex:CollectionManifestation1/geut rdamo:P30460 ex:Item //(1, 2, etc.) ex:CollectionManifestation1/geut rdamd:P30277 "Collection of Candler School of Theology" ex:CollectionManifestation/geut rdamo:P30135 ex:CollectionWork/geut ex:CollectionWork/geut rdawd:P10329 "Collection of Candler School of Theology" ex:CollectionWork/geut rdawo/P10631 ex:CollectorCorporateBody1/geut ex:CollectorCorporateBody1/geut rdaa:P50383rdaa ex:Nomen1 ex:Nomen1 rdand:P80033 “GEU-T” ex:Nomen1 rdano:P80069 "https://www.loc.gov/marc/organizations/" |
Beta Was this translation helpful? Give feedback.
-
|
Replying to @GordonDunsire's last post - I'm concerned about the different organization levels in the names given for the Organization Codes and Cultural Heritage organization lists, and the names we would expect to be the "holding institution" (library, archive, etc.). They may be the same in some cases but for e.g. Emory's GEU-T, the name given is Candler School of Theology (Org Codes) or Emory University. Candler School of Theology (Cultural Heritage), rather than the Pitts Theology Library (which is under the Candler School of Theology organizationally). These are distinct, if organizationally linked, corporate bodies and I think it's risky to conflate or confuse them. I realized this morning that I forgot, in my example, to name the minted collection work and manifestation. We were going to give it a "Collection of..." string label. Should it be "Collection of Candler School of Theology" (derived from the MARC code list)? The collection is actually the library collection of the Pitts Theology Library (the school itself may have other collections somewhere but never mind). I don't think we can derive that fact from the data available to us. |
Beta Was this translation helpful? Give feedback.
-
|
I have added the appellations to the collection manifestation and work entities in the example. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Another draft for review/revision/comment - statement to be included in documentation of the mapping /conversion : For converting MARC data related to item entities, RDA does not define an intermediate entity indicating "the holdings" of a manifestation by an institution, i.e. the subset of an institution's collected holdings associated with a manifestation. Therefore this mapping generates a different item URI for each field containing, for example, a $5 indicating item-specific data, in order to establish a relationship with a manifestation and apply notes or other relations according to the field content. The result may easily have redundant item IRIs when only a single item exists, since practice at many institutions results in $5 in multiple fields in a description of, for example, rare materials or archives. Data in MARC is not sufficient to identify specific items or relate item data from different parts of the record with specific generated IRI identifiers. Item data may thus require manual reconciliation after conversion, depending on the institution's needs. |
Beta Was this translation helpful? Give feedback.
-
|
I think this is fine, but ... Strictly speaking, RDA does accommodate the 'manifestation holdings' of a specific library as a sub-collection manifestation. The subset of items that exemplify a manifestation and that are held by one organization form a fine-grained collection manifestation and work that can be recorded like this: Manifestation1 - has exemplar of manifestation - Item1 . Item 1 - is holding of - CollectionManifestation1 . CollectionManifestation1 - is subcollection of - OrganizationCollectionManifestation . CollectionManifestation1 is the collection of holdings of a single manifestation by a single organization. This approach may be useful outside of the scope of this project, for mapping non-MARC propriety (e.g. Alma) 'holdings' records. Also, in practice, most transform users would wish to add their local item identifiers in due course, including barcodes, ms, and archival identifiers, and this is when reconciliation of multiple IRIs can take place |
Beta Was this translation helpful? Give feedback.
-
|
Ok, dropping that sentence. |
Beta Was this translation helpful? Give feedback.
-
|
Example MARC mapping (from Emory record MMSID (001) 990000266450302486) 500 a| Initial letter E illuminated in blue, green, gold, purple and black; elaborately decorated at foot of page with coat of arms. Initials supplied in red. 5| GEU-T Resulting LRM/RDA/RDF Triples: ex:Manifestation1[MintIRI] rdamo:P30103[has exemplar of manifestation] ex:Item1[MintIRI] |
Beta Was this translation helpful? Give feedback.
-
|
This is linked to in the Decision on $5 as of 9/15/2022 |
Beta Was this translation helpful? Give feedback.
-
|
At a loss for how to map this. 49 items for each of the institutions in $5? https://www.loc.gov/marc/bibliographic/bd538.html 538 ## $3 v.1-49(1927-1975) $a Master and use copy. Digital Master created according to Benchmark for Faithful Digital Reproductions of Monographs and Serials, Version 1. Digital Library Federation, December 2002. $u http://www.diglib.org/standards/bmarkfin.htm $5 NIC $5 ICU |
Beta Was this translation helpful? Give feedback.
-
|
Odd example for the 538. It doesn't really tell us anything about the system details required to view or use the resource. The previous example in the MARC format seems more typical:
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Zhuo Pan ***@***.***>
Sent: Wednesday, October 12, 2022 5:03 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
At a loss for how to map this. 49 items for each of the institutions in $5?
https://www.loc.gov/marc/bibliographic/bd538.html [loc.gov]<https://urldefense.com/v3/__https://www.loc.gov/marc/bibliographic/bd538.html__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvuS0YG3M$>
538 ## $3 v.1-49(1927-1975) $a Master and use copy. Digital Master created according to Benchmark for Faithful Digital Reproductions of Monographs and Serials, Version 1. Digital Library Federation, December 2002. $u http://www.diglib.org/standards/bmarkfin.htm [diglib.org]<https://urldefense.com/v3/__http://www.diglib.org/standards/bmarkfin.htm__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvXDxIDfQ$> $5 NIC $5 ICU
—
Reply to this email directly, view it on GitHub [github.com]<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/361*discussioncomment-3865113__;Iw!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvMSDqfEg$>, or unsubscribe [github.com]<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVBYKA4BVKFVXRZSCLV3WC5GWPANCNFSM5WQBZS4Q__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvsifVRRU$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Damn, I didn't mean to send this message yet, it wasn't complete. Let's try again
Odd example for the 538. It doesn't really tell us anything about the system details required to view or use the resource. The previous example in the MARC format seems more typical:
538 ##$31-39(1927-1965)$aFiles for the images of individual pages are encoded in Aldus/Microsoft TIFF Version 6.0 using facsimile- compatible CCITT Group 4 compression.$5NIC
But even with this example, your question is still applicable. The info is $3 is sort of holdings information for the resource as a whole. The 538 would typically be manifestation level information I think. I don't know if you'd have to have 39 separate item statements, or somewhere in manifestation description, or if this should just be treated as outside of WEMI.
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Adam L Schiff ***@***.***>
Sent: Thursday, October 13, 2022 12:18 PM
To: uwlib-cams/MARC2RDA ***@***.***>; uwlib-cams/MARC2RDA ***@***.***>
Cc: Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
Odd example for the 538. It doesn't really tell us anything about the system details required to view or use the resource. The previous example in the MARC format seems more typical:
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
________________________________
From: Zhuo Pan ***@***.***>
Sent: Wednesday, October 12, 2022 5:03 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
At a loss for how to map this. 49 items for each of the institutions in $5?
https://www.loc.gov/marc/bibliographic/bd538.html [loc.gov]<https://urldefense.com/v3/__https://www.loc.gov/marc/bibliographic/bd538.html__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvuS0YG3M$>
538 ## $3 v.1-49(1927-1975) $a Master and use copy. Digital Master created according to Benchmark for Faithful Digital Reproductions of Monographs and Serials, Version 1. Digital Library Federation, December 2002. $u http://www.diglib.org/standards/bmarkfin.htm [diglib.org]<https://urldefense.com/v3/__http://www.diglib.org/standards/bmarkfin.htm__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvXDxIDfQ$> $5 NIC $5 ICU
—
Reply to this email directly, view it on GitHub [github.com]<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/361*discussioncomment-3865113__;Iw!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvMSDqfEg$>, or unsubscribe [github.com]<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVBYKA4BVKFVXRZSCLV3WC5GWPANCNFSM5WQBZS4Q__;!!K-Hz7m0Vt54!hBXquIY_WSkZuzgK_bXYGT2-WlcrwFNVeoPm5fgWY5mhYftnsLcp0zAWCeQsqkUSw-yzdWcaEpJQVbOvsifVRRU$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
I wonder if $3 in both MARC 21 manual examples is consistent and refers to some, but not all, of the manifestation. The manifestations are presumably digitized back-runs of a diachronic work, and the implication of the encoding is that it is not the complete(d) back-run. If so, the multiple institutions in $5 are consistent with 'holding' collections. There is one item for each institution; each item is a distinct copy of the digital file that is the manifestation. The decision to mint such items is the same as for other instances of $5. Each minted item will consist of a common statement ( has manifestation exemplified ), a specific identifier statement ( has identifier for item "stringify (this item IRI)"), and a specific holding statement ( is holding of ) - but no other information. The $3 data belongs to the manifestation, and applies to all of its items. The only characteristic that distinguishes the items is the holding institution, and that statement may no longer be valid. Is it worth minting any of these items? |
Beta Was this translation helpful? Give feedback.
-
|
The problem is when you have $3 and $5 together in a field. $3 could be "booklet" or "v. 1-3"; the former implies only one item for the booklet while the latter might suggest 3 items, one for each volume. Given that $3 is uncontrolled we could only mint one item for each $5. And if we only append the string "Applies to: XXX" to "Note on manifestation", then we are essentially only saying that the $5 institution has an item of the manifestation and almost nothing else about that item. |
Beta Was this translation helpful? Give feedback.
-
|
This might cause us to question our assumptions leading to creating items and mapping note text to an item description instead of a note on manifestation. But we didn't find a good alternative then... Because we're working from example tags without the context of the record, it's harder to get the story right, but for the 2 538's in the MARC documentation, I imagine that what is being described is a digital file that was created by and for a particular institution (hence the $5). Why else would technical details be institution-specific? Certainly we wouldn't encounter a record that describes both a physical manifestation and an electronic one, would we? I thought that wasn't kosher, or have the rules changed? If for this field we are to create an item in order to indicate the owner of the item in order to get the "institution to which field applies" relationship, the note needs to become a Note on item, doesn't it? We already have said we will have to caution people that a single item may not represent the actuality of their holdings and they might want to revise these statements. I think of this item conceptually as an "item representative of holdings" but that's just in my head. If you wanted to leave it a Note on manifestation for this field, I don't think we could come up with another way to indicate that the note applies to a particular institution than to add something more (besides content of the subfield 3) to the end of the note (like, Applies to holdings of: Cornell University (NIC) and University of Chicago (ICU)). The fact there are two $5's in that second example is what's mind blowing to me. Does one possess the digital master and the other possess the use copy? Or do they share a server as part of some project... ? |
Beta Was this translation helpful? Give feedback.
-
|
I think Laura's last suggestion is the most sensible solution for the $3/$5 occurring together: If you wanted to leave it a Note on manifestation for this field, I don't think we could come up with another way to indicate that the note applies to a particular institution than to add something more (besides content of the subfield 3) to the end of the note (like, Applies to holdings of: Cornell University (NIC) and University of Chicago (ICU)). The fact there are two $5's in that second example is what's mind blowing to me. Does one possess the digital master and the other possess the use copy? Or do they share a server as part of some project... ? Append to note on manifestation: Applies to: [$3 content] at [institution(s) in $5] I agree with Zhuo's thinking that it is not useful to mint items in these cases where something maps to a note on a manifestation...we have nothing to say about these items. Is this where others have landed here? |
Beta Was this translation helpful? Give feedback.
-
|
Just to capture here: why I think decision on $5 mapping should be a tag-specific judgment call because of the ambiguity of its definition: MARC code of the institution or organization that holds the copy to which the data in the field applies. Data in the field may not apply to the universal description of the item or may apply universally to the item but be of interest only to the location cited. In case of the example 538 which also has $3, my judgment says this statement might apply universally to something (the word item is used here but I think manifestation is closer to what the meaning has to be, does a single copy have a universe?), or might not, we don't know, but it probably has the $5 because the note is of interest to that institution. Hence, use boilerplate to add to the manifestation note rather than minting an item and making it an item note. |
Beta Was this translation helpful? Give feedback.
-
|
I think I understand what you mean now, please correct me if I'm getting it wrong: The $5 may indicate copy-specific information, or information that applies to the entire manifestation but is only of interest to the institution in the $5. There is not a way to distinguish between these two scenarios without looking at individual notes. So, we need to decide between our previous decision to mint items for $5 or mapping every $5 in a similar way that we have done when a $3 is involved, with boilerplate. Since not everything with a $5 is a note, and sometimes may be very item-specific depending on the tag (thinking of a 700 with $e "previous owner" and a $5 for institution), we need to decide whether to use the boilerplate approach or the item-minting approach on a case-by-case basis. Since Zhuo is almost done with the transform on $5, maybe we ought to circle back to this once he has finished? I will put scheduling for this on next week's agenda (5 minutes, "when should we talk about this", not "discuss in depth now") |
Beta Was this translation helpful? Give feedback.
-
|
I can't say that I've ever seen any use of the $5 for the latter part of the definition, and I seriously doubt that anyone uses it that way. But I haven't seen every MARC record with a $5!
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Crystal Yragui (Clements) ***@***.***>
Sent: Wednesday, November 16, 2022 10:25 AM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
I think I understand what you mean now, please correct me if I'm getting it wrong: The $5 may indicate copy-specific information, or information that applies to the entire manifestation but is only of interest to the institution in the $5. There is not a way to distinguish between these two scenarios without looking at individual notes. So, we need to decide between our previous decision to mint items for $5 or mapping every $5 in a similar way that we have done when a $3 is involved, with boilerplate. Since not everything with a $5 is a note, and sometimes may be very item-specific depending on the tag (thinking of a 700 with $e "previous owner" and a $5 for institution), we need to decide whether to use the boilerplate approach or the item-minting approach on a case-by-case basis.
Since Zhuo is almost done with the transform on $5, maybe we ought to circle back to this once he has finished? I will put scheduling for this on next week's agenda (5 minutes, "when should we talk about this", not "discuss in depth now")
—
Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/361*discussioncomment-4159980__;Iw!!K-Hz7m0Vt54!kPnoSmgxTPxhHQV1wG0bgKZIAomOGXE5q23YIiNaD2eh89hIk7D62CmHiBU338VcdYyeUInEdIEaAa6p5VCLxXg$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVBZ75GKDMJAEOZMWGF3WIURH5ANCNFSM5WQBZS4Q__;!!K-Hz7m0Vt54!kPnoSmgxTPxhHQV1wG0bgKZIAomOGXE5q23YIiNaD2eh89hIk7D62CmHiBU338VcdYyeUInEdIEaAa6pEK1odLs$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
I am actually done with $5 preprocessing and 500 for now and I am waiting for mapping decisions. The preprocessed collection instances for Cultural Heritage Organizations can be found at The test output for 500 with $5 If there are multiple $5's in a record, the item IRIs end with ite1, ite2, ite3... which are based on the positions of the $5's in the record. The caveat is that when the record is updated and new $5's added, the relative positions of $5's may change, and when you rerun the transform the item IRIs change. Needs to be addressed in the future. There's also the issue of invalid codes in $5. I've created some fake data just to see what the result looks like when $3 and $5 appear together. Output: |
Beta Was this translation helpful? Give feedback.
-
|
Thank you Zhuo! I'll bring this up at our next meeting |
Beta Was this translation helpful? Give feedback.
-
|
attn: @lake44me |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI Sorry I missed the last meeting. I was on vacation, meant to come, but got embroiled in family holiday stuff and forgot to tell you. Looks like you were discussing my favorite subfield. Lo, the discussion when $5 was proposed to be added to tags 533 and 538 is online: https://www.loc.gov/marc/bibliographic/ecbdcntf.html . I had a subiliminal notion about this and it turns out the $5 on these tags was added to support the Registry of Digital Masters program, particularly for electronic serials which follow a provider-neutral policy of covering all electronic versions in one record. I think the examples in the tag definitions are probably taken from that program. From my reading of it, there's a clear indication that presence of $5 in these tags indicates that the content is copy-specific (a digital master), and so I take back my waffling on how to handle it. I think it would be better to mint a copy and follow the same process for these that we do for other note tags, identifying the institution in $5 as holding the copy described in the note. |
Beta Was this translation helpful? Give feedback.
-
|
What to do when it's clear that $5 isn't paired with item information? e.g. 588 Source of Description Note. From CONSER editing guide (https://www.loc.gov/aba/pcc/conser/CEG/588.doc):
Also note that the $5 codes in 588 are not from id.loc.gov Cultural Heritage Organizations, but OCLC symbols. |
Beta Was this translation helpful? Give feedback.
-
|
@cspayne found another field with $5 that does not indicate copy-specific information. Field 526. I propose that we add boilerplate in these cases, appending "Applies to: [Institution name associated with code in $5]" in cases like these. @lake44me @AdamSchiff @GordonDunsire is this a workable solution for you? We've let Zhuo's question hang since May, and the 588 mapping still has status "?" for $5. Unless any of you strongly object, I can update the 588 and Cypress will use this approach in mapping the 526, and I will record some sort of decision about it in our index. Eh? |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI do we want to try and lookup the institution name associated with the $5 code or just say "Applies to: $5 code"? The transform can do a lookup for $5 codes from id.loc.gov, but I'm not sure if there's a way to access a list of OCLC symbols. |
Beta Was this translation helpful? Give feedback.
-
|
Crystal, just a guess but, it would seem to me that for 588 the institution symbol in $5 would indicate that that institution is responsible for making the statement about the source of description. Description of what? Closest match would be manifestation.
We don't have an entity for this description. Yet. It's sort of metadata about the metadata we are mapping...
Would this be kosher?
http://rdaregistry.info/Elements/m/P30462 (Manifestation described with metadata by) - minted IRI for a metadata work
MW IRI Title of work "Source of Description Note"
MW IRI http://rdaregistry.info/Elements/w/P10402 (Note on metadata work) "(contents of 588 subfield a)"
MW IRI http://rdaregistry.info/Elements/w/P10531 (Creator corporate body of work) "either the symbol or a look up of the institution represented by the symbol"
Well, maybe it would provide food for discussion by Those Who Know ..
566 (Study Program Information Note) is different. I'd be ok with just mapping it to Note on Manifestation with a prefix, concatenating and labeling the contents of the subfields, and appending something about the $5 institution, "applies to" or "at institution" would be fine. However, more challenges lurk in this tag! There are subfields for "public note" and "nonpublic note". How are you going to handle the nonpublic ones? Reify in some way and assign category of work "nonpublic" like we did in 561 for the nonpublic indicator fields?
Enjoy your holiday, don't think about this too much, especially with eggnog...
Laura
From: Crystal Yragui (Clements) ***@***.***>
Sent: Thursday, December 21, 2023 4:31 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Laura Akerman ***@***.***>; Mention ***@***.***>
Subject: [External] Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
@cspayne<https://github.com/cspayne> found another field with $5 that does not indicate copy-specific information. Field 526. I propose that we add boilerplate in these cases, appending "Applies to: [Institution name associated with code in $5]" in cases like these. @lake44me<https://github.com/lake44me> @AdamSchiff<https://github.com/AdamSchiff> @GordonDunsire<https://github.com/GordonDunsire> is this a workable solution for you? We've let Zhuo's question hang since May, and the 588 mapping still has status "?" for $5. Unless any of you strongly object, I can update the 588 and Cypress will use this approach in mapping the 526, and I will record some sort of decision about it in our index. Eh?
-
Reply to this email directly, view it on GitHub<#361 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAKBM23KAGFVZJCYQ56SECLYKSTAHAVCNFSM5WQBZS42U5DIOJSWCZC7NNSXTOKENFZWG5LTONUW63SDN5WW2ZLOOQ5TOOJSGIZDQNY>.
You are receiving this because you were mentioned.Message ID: ***@***.******@***.***>>
|
Beta Was this translation helpful? Give feedback.
-
|
Organization codes used in $5 are available from id.loc.gov at https://id.loc.gov/vocabulary/organizations.html. Here's UW Libraries: http://id.loc.gov/vocabulary/organizations/wau Our actual MARC code is WaU, but that gets normalized to all lower case in the URI.
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Cypress ***@***.***>
Sent: Thursday, July 4, 2024 1:46 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
@CECSpecialistI<https://urldefense.com/v3/__https://github.com/CECSpecialistI__;!!K-Hz7m0Vt54!nbBvgtQ66FT4f6XyP-l4pFT6bxlXSqGwm2ih9ERiTU6s7k1qFwj7ikz4gEAphKVjLEl0-sEzc_k8xwNv2CL7y_w$> do we want to try and lookup the institution name associated with the $5 code or just say "Applies to: $5 code"? The transform can do a lookup for $5 codes from id.loc.gov, but I'm not sure if there's a way to access a list of OCLC symbols.
—
Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/361*discussioncomment-9962281__;Iw!!K-Hz7m0Vt54!nbBvgtQ66FT4f6XyP-l4pFT6bxlXSqGwm2ih9ERiTU6s7k1qFwj7ikz4gEAphKVjLEl0-sEzc_k8xwNvBiwiLQE$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVB4CAN5DO5RMVO5UFLDZKWYATAVCNFSM5WQBZS42U5DIOJSWCZC7NNSXTOKENFZWG5LTONUW63SDN5WW2ZLOOQ5TSOJWGIZDQMI__;!!K-Hz7m0Vt54!nbBvgtQ66FT4f6XyP-l4pFT6bxlXSqGwm2ih9ERiTU6s7k1qFwj7ikz4gEAphKVjLEl0-sEzc_k8xwNvYfjlCvg$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
Thank you @AdamSchiff ! I think the solution for now should be to lookup and retrieve institution names from id.loc.gov using the $5 code, and in the (hopefully rare) case that the code cannot be reconciled, to use the code value instead of the institution name. This is solely for when we are including $5 as part of a note, not when we are minting an item as part of a collection based on $5. |
Beta Was this translation helpful? Give feedback.
-
|
I first thought that you shouldn't need to even do a lookup since there's a pattern. Lowercase the code in $5 and append it to http://id.loc.gov/vocabulary/organizations/ However, this will only work for U.S. and some other libraries elsewhere. "While this list of organizations focuses on US institutions, with over 30,000 defined, it also includes codes for institutions in other countries that have requested them. However, MARC codes are not assigned for institutions for Canada, Germany, or the United Kingdom unless the institution is a branch of a US institution. The list contains over 41,000 entries." So maybe a lookup would be best, since the machine can't tell if the code is for a U.S. institution or for a library in Canada, Germany, or U.K. which probably doesn't have a URI.
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Cypress ***@***.***>
Sent: Friday, July 5, 2024 3:19 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] $5 (Discussion #361)
Thank you @AdamSchiff<https://urldefense.com/v3/__https://github.com/AdamSchiff__;!!K-Hz7m0Vt54!imug2GxSbZEhITlqxDNuAVFGxTAprALxcDipxzdW3xG75WLASzbO95r1CuEuq3WMRFj8oriTy6igv80rIp-PgLk$> !
I think the solution for now should be to lookup and retrieve institution names from id.loc.gov using the $5 code, and in the (hopefully rare) case that the code cannot be reconciled, to use the code value instead of the institution name. This is solely for when we are including $5 as part of a note, not when we are minting an item as part of a collection based on $5.
—
Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/361*discussioncomment-9971902__;Iw!!K-Hz7m0Vt54!imug2GxSbZEhITlqxDNuAVFGxTAprALxcDipxzdW3xG75WLASzbO95r1CuEuq3WMRFj8oriTy6igv80rehj41Jk$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVB4ILARMOBAT73TGF43ZK4LXZAVCNFSM5WQBZS42U5DIOJSWCZC7NNSXTOKENFZWG5LTONUW63SDN5WW2ZLOOQ5TSOJXGE4TAMQ__;!!K-Hz7m0Vt54!imug2GxSbZEhITlqxDNuAVFGxTAprALxcDipxzdW3xG75WLASzbO95r1CuEuq3WMRFj8oriTy6igv80rZZReGSE$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
@AdamSchiff @lake44me @CECSpecialistI |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I think we should be minting a new item for each $5 subfield, even within the same field. |
Beta Was this translation helpful? Give feedback.
-
|
I agree! |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
$5 previously mentioned in:
Reification in the mapping #351
Beta Was this translation helpful? Give feedback.
All reactions