Publications list -- link entries to full bib record in blacklight.

KraigBinkowski commented 4 years ago

Within the publications list display -- can these individual entries be made to link directly to the full bib records in Blacklight?

flapka commented 4 years ago

@KraigBinkowski I like this suggestion. When RBM revises its procedures for publications information (to mimic what's in place already for art collections) this linking functionality would be desirable for us too.

yulgit1 commented 4 years ago

Does TMS have bibids that can be ported into lido and indexed along with each citation?

Or is there another way to make this link? It is somewhat possible using solr, but not advisable, as would quickly drag down performance, and would be difficult to formulate a search that finds 1 and only 1 associated record.

flapka commented 4 years ago

@yulgit1 @KraigBinkowski Speaking with only a novice understanding TMS Bibliography module: I echo Eric's suggestion. It seems like it'd be really useful to have the Orbis bib ID as part of the data in TMS. Could it be recorded as an Alternate Number, of type YaleBibID (it won't be Orbis forever). With that piece of information expressed (in the LIDO), it seems like it should trivial to make the links.

If RB revises its practice to mimic the art collections publications format, I think we'd want to rely on (put in MARC) the Orbis ID to make such connections.

KraigBinkowski commented 4 years ago

right now the bibids are not part of the information added to the TMS bib module. brings up a couple of questions - how does this look for articles vs books and is there a way to mass import bibIDs into the bib module without adding them one by one manually (for existing records)?

flapka commented 4 years ago

On the mass import, that seems doable. If TMS could export a table for the Alternate numbers, I think we could do a join query with a table for items in REF, using the call number as the linking property, adding the bib ID to the TMS table.

I might even be able to do this myself, if the TMS table could be provided.

edgartdata commented 4 years ago

@flapka @KraigBinkowski Alternatively, it would be very easy to export bibliographic records IDs out of TMS. The reason I am mentioning this is that I don't think it matters where the joint between TMS and MARC is made, right? Otherwise, using Alternate Number in the TMS Bib Module should work, although we would need to check with David that this could be a bulk ingest.

edgartdata commented 4 years ago

@yulgit1 @KraigBinkowski @flapka returning to this issue as it is now in scope for LUX PoC (you are welcome!). From the convo above, is the next step for me to check with David Parsell if it is possible to bulk ingest OrbisBibIDs in TMS predicated on the titles of the citations? Titles might not be distinct enough for matches. Then what other identifiers/metadata elements could we use?

KraigBinkowski commented 4 years ago

That's great news! Hmmm - the only truly unique field in the record is the call number, which by definition is used once per title. Or is there the option of matching bibids to a combo of title and author? and then having a report of bibids that could not be matched to resolve manually?

flapka commented 4 years ago

@KraigBinkowski @edgartdata @yulgit1

I 2nd Kraig's suggestion. Something like:

Step 1: Attempt to execute a union query (TMS bib module / Orbis data) based on call numbers in each data set.
Step 2: Manually review and edit data where the union query was not successful
Step 3: Re-ingest data to TMS with Orbis bib IDs!

I'd be happy to provide a table of Ref's Orbis data, if that'd be helpful. Do citations also reference Orbis records outside of Ref?

KraigBinkowski commented 4 years ago

A small subset of records are non Ref Orbis records -- probably less than 5%. These would be Haas and/or Sterling Lib. -- Perhaps easiest to segregate these and add bibiIDs manually?

flapka commented 4 years ago

@KraigBinkowski Yes, that sounds reasonable. Especially as a good working from home project, perhaps.

edgartdata commented 4 years ago

This is the entire list of bib citations in TMS (at least the ones that have a title) and 3,894 out of these 4,060 have a LC Call Number. TMS has other types of identifiers for some of these items:

Call Number (is that different from LC Call Number?)
Class Number
Control Number
Dewey Number
Government Number
ISNB
ISSN
LC Call Number
LC Control Number
OCLC Number

Unfortunately the TMS list view mode that exports to Excel only grabs the first identifier listed in the TMS Alternate Number field. I can generate smaller groups per number type if that's helpful though and see if there is a way to get all helpful identifiers into one document for this data reconciliation task.

TMS Bib records September 3 2020.xlsx

yulgit1 commented 4 years ago

Could the bibid be exposed in the LIDO something like:

34 4226255

flapka commented 4 years ago

@KraigBinkowski @yulgit1 Reviewing the table that @edgartdata created and comparing it to a table of Ref library data reveals a couple of points:

There are subtle differences in call number syntax that are likely to make automated joins on call number unsuccessful. For example: TMS has "N12 W35 +A1 OVERSIZE (YCBA)" where Voyager has "N12.W35 A1 (LC) Oversize)" for this item.
On the other hand, it looks like the number of unique Orbis titles represented in TMS citations is much smaller than one might think. I did a grouping represented in the Excel file loaded herewith. If I crunched correctly, it appears that (from the 4000+ citations) there are only 115 unique ReferenceIDs and 100 unique call numbers. Unless I've made a mistake in the crunching, it looks like a manual addition of Voyager IDs to this table (of 115 entries) might do the trick?

citations_grouped.xlsx

KraigBinkowski commented 4 years ago

most of those identifiers are only present in the records we imported to TMS via Z39.50 protocol. We stopped doing that quite a while ago so they are only present in a smaller portion of the citations. All of the actual call number data should be in the "LC call number" field -- if data is in the "Call Number" field its probably a mistake.

KraigBinkowski commented 4 years ago

If what you are saying is that there are only 115 unique books represented by the 4000 citations than I would say there has to be a mistake. I don't have the exact numbers at hand but that really can't be the case as we have been adding books and articles for over 10 years. The vast majority of these additions are small numbers of citations taken from books and not whole books containing a large amount of citations. If we can run a report on the titles in the bib module, that would give us an accurate count.

flapka commented 4 years ago

@KraigBinkowski I too found the idea of only 115 unique entries surprising, so I wonder if something is amiss.

The spreadsheet that @edgartdata shared has 4060 rows, of which 93% come from just four (!) titles:

ReferenceID	Title	Alternate Number	CountOfAlternate Number
323	A Concise Catalogue of Paintings in the Yale Center for British Art	N590.2 A83 (YCBA)	3021
1143	A Catalogue of the Paintings, Drawings, and Etchings of Philip Mercier	N12 W35 +A1 OVERSIZE (YCBA)	424
3977	A Decade of Gifts and Acquisitions	V2724 (YCBA)	182
1732	A biographical dictionary of sculptors in Britain, 1660-1851	NB496 B56 2009 (YCBA) (Wall Shelf 3)	171

Is it possible that the TMS export didn't snag all the entries?

KraigBinkowski commented 4 years ago

Yeah something is amiss - from our latest statistics (last annual report) in this one year, staff added 513 citations and 441 book and article titles to TMS, I wrote that this "ongoing effort has added over 14,000 bibliographic citations to 4,382 objects in the Center’s collection since the project began in 2009. " So, the number should be closer to 14000 citations than 4000. These figures are based on several reports that David devised to try and quantify the process.

edgartdata commented 4 years ago

I think you are right. I did not get all the TMS bib records out because all searches I do not on other fields return the same number of results, i.e. 4,060. Let me work on this and get back to you.

edgartdata commented 4 years ago

@KraigBinkowski @flapka

An update on this issue: David Parsell checked the TMS SQL tables and he finds the same number of TMS Bibliography Module records as I do, i.e. 4,060. We wondered for a moment if TMS might have "lost" records but a copy of the TMS database from May 4 2020 also has 4060 records so I think we can be pretty confident that this number is accurate.

What I think happened is that the TMS export mechanism into Excel had a problem. It seems that the process timed out somehow and the Excel list got populated with many duplicates records instead of real records. Unfortunately this export process is still not working today.

I have asked David for a list (from SQL) of all bib records that have LC Call Numbers and another one with all Bib Module records.

edgartdata commented 4 years ago

@edgartdata to identify the TMS Bib records that do not have LC Call Numbers.

edgartdata commented 3 years ago

@KraigBinkowski @flapka I have checked all the reports that we currently have available in the TMS Bib Module to export bib citations out of TMS and all the reports contains many duplicate unfortunately.

To get this moving, I queried TMS for Bib records that have a title (I figured 99% of TMS bibliographic citation records have a title), exported the results to Excel via a TMS report and then took out all the duplicated rows out of the Excel (based on the TMS ReferenceID). The result is this list of 4,832 TMS Bib Module records with 3 columns: TMS Bib Module ReferenceID, Title, and LC Call Number:

YCBA TMS Bib Module records with titles 1-29-2021.xlsx

I realize that my list might not be totally accurate in terms of number of Bib records but I figured it is close enough to get us started.

As we talked about, I hope this list helps you match the TMS Bib Module records in this list with their Orbis BibID, so that we can make actionable the titles that appear in the Publications section of RBM or P&D/P&S BlackLight records and have these titles link to their corresponding more complete Reference Library BL records.

KraigBinkowski commented 3 years ago

Thanks Emmanuelle - this number sounds like it could be close to the mark - given that many titles have several different objects cited. Francis and I can take a look and see if your report gives us what we need to be able to come up with a matching plan. Is it possible to include more fields to the spread sheet if we need them (thinking of author)?

Thanks very much for compiling this,

Kraig

Kraig Binkowski Chief Librarian Reference Library and Archives

Yale Center for British Art

1080 Chapel Street

P.O. Box 208280

New Haven, CT 06520-8280

+1 203 432 2846 | +1 203 415 1192 mobile

britishart.yale.eduhttp://britishart.yale.edu/

Explore the Center's exhibitionshttps://britishart.yale.edu/exhibitions online. Connect on Facebookhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.facebook.com%2FYaleBritishArt%2F&data=02%7C01%7Clyn.rose%40yale.edu%7C02164475b8f54640e50d08d79f5be394%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637153089312502268&sdata=cWFWX9dZKjaC9SiTfDHkWQmsW0fv5Y%2B9aaSfG%2FWSo%2BA%3D&reserved=0, Twitterhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Ftwitter.com%2FYaleBritishArt&data=02%7C01%7Clyn.rose%40yale.edu%7C02164475b8f54640e50d08d79f5be394%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637153089312512260&sdata=IFyktJi9i3QXR94LWitH5svX0D%2F9pQ8Bb4Vjjm8BWGI%3D&reserved=0, Instagramhttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.instagram.com%2Fyalebritishart%2F&data=02%7C01%7Clyn.rose%40yale.edu%7C02164475b8f54640e50d08d79f5be394%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637153089312512260&sdata=7hZ4%2BD6l1XUl13qTvjpB56lQ8hVH4VvnrhhR3lnKG4w%3D&reserved=0, and YouTubehttps://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.youtube.com%2Fyalebritishart&data=02%7C01%7Clyn.rose%40yale.edu%7C02164475b8f54640e50d08d79f5be394%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637153089312522255&sdata=RQmjS7zAcuAEdlGLt55LVMHp7gds5pqmys3g2PPxlWk%3D&reserved=0 @yalebritishart #YCBA

From: Emmanuelle Delmas-Glass notifications@github.com Sent: Friday, January 29, 2021 11:51 AM To: ycba-cia/blacklight-collections2 blacklight-collections2@noreply.github.com Cc: Binkowski, Kraig kraig.binkowski@yale.edu; Mention mention@noreply.github.com Subject: Re: [ycba-cia/blacklight-collections2] Publications list -- link entries to full bib record in blacklight. (#211)

@KraigBinkowskihttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FKraigBinkowski&data=04%7C01%7Ckraig.binkowski%40yale.edu%7Caf26381250e74f3df86308d8c476077f%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637475358505808156%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=Bj0L4Cp9OZd3osyycv72uE%2FyGWIwFzSy5Q89sSMEwYw%3D&reserved=0 @flapkahttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fflapka&data=04%7C01%7Ckraig.binkowski%40yale.edu%7Caf26381250e74f3df86308d8c476077f%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637475358505818150%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=xQwBEnfEXZQwazC6MF7gqH3EbVaAT6E4q3SjwTdezPI%3D&reserved=0 I have checked all the reports that we currently have available in the TMS Bib Module to export bib citations out of TMS and all the reports contains many duplicate unfortunately.

To get this moving, I queried TMS for Bib records that have a title (I figured 99% of TMS bibliographic citation records have a title), exported the results to Excel via a TMS report and then took out all the duplicated rows out of the Excel (based on the TMS ReferenceID). The result is this list of 4,832 TMS Bib Module records with 3 columns: TMS Bib Module ReferenceID, Title, and LC Call Number:

YCBA TMS Bib Module records with titles 1-29-2021.xlsxhttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fycba-cia%2Fblacklight-collections2%2Ffiles%2F5895069%2FYCBA.TMS.Bib.Module.records.with.titles.1-29-2021.xlsx&data=04%7C01%7Ckraig.binkowski%40yale.edu%7Caf26381250e74f3df86308d8c476077f%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637475358505818150%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=Chijlj6CABDsXjSoTL3PKiThhM27V%2BQig7xLcBICFHQ%3D&reserved=0

I realize that my list might not be totally accurate in terms of number of Bib records but I figured it is close enough to get us started.

As we talked about, I hope this list helps you match the TMS Bib Module records in this list with their Orbis BibID, so that we can make actionable the titles that appear in the Publications section of RBM or P&D/P&S BlackLight records and have these titles link to their corresponding more complete Reference Library BL records.

- You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fycba-cia%2Fblacklight-collections2%2Fissues%2F211%23issuecomment-769920486&data=04%7C01%7Ckraig.binkowski%40yale.edu%7Caf26381250e74f3df86308d8c476077f%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637475358505828143%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=%2FO13DQGZF1nupMYo4oJGuuv%2Ft7k1wPeih24kV1tiftI%3D&reserved=0, or unsubscribehttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FALGDYPUAD2UXFQMGOUGVGALS4LROPANCNFSM4NSBXYHQ&data=04%7C01%7Ckraig.binkowski%40yale.edu%7Caf26381250e74f3df86308d8c476077f%7Cdd8cbebb21394df8b4114e3e87abeb5c%7C0%7C0%7C637475358505828143%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=u33RSRiThWLjgI1vrhXdnxESjNN5ASqh1rC0rupaMlM%3D&reserved=0.

KraigBinkowski commented 3 years ago

Thanks Emmanuelle – this number sounds like it could be close to the mark – given that many titles have several different objects cited. Francis and I can take a look and see if your report gives us what we need to be able to come up with a matching plan. Is it possible to include more fields to the spread sheet if we need them (thinking of author)?

flapka commented 3 years ago

With the hope of doing semi-automated matching, is it possible to export a table that includes the ISBN numbers and the "Control Number" whenever a record includes these numbers? Matching on call number can be tricky, because of variations in syntax. ISBN numbers and control numbers (which are equal to BIB ID?), where present, might have fewer issues of complicated syntax.

yulgit1 commented 3 years ago

As a TMS learning exercise I took a stab at this. The number of objects is a little different 3832 in EDG's vs 4146. @flapka let me know if this has all necessary and correct fields.

bib-report-erj-jan29-2021.xlsx

KraigBinkowski commented 3 years ago

Thanks Eric - this looks great. I wonder what accounts for the 300 or so difference in titles? Francis, can we meet next week and come up with a plan for fields to use to automate the process of importing the bibids? This looks like a promising start.

flapka commented 3 years ago

@yulgit1 For some of the items in your report, the value for ISBN looks corrupted. For example (the third column):

1298	Gardens and cultural change	9.7808840233e+12	SB470.7 .G37 2007 (YCBA)	NULL	8152450

In TMS that ISBN value is 9780884023302. Is it corrupted by a data type version? Regardless, could you generate the report again in a way that addresses this issue? Thanks!

edgartdata commented 3 years ago

@flapka @KraigBinkowski @yulgit1 Could it be that the ISBN value for Gardens and cultural change is corrupted is because that title is not linked to any TMS object records?

yulgit1 commented 3 years ago

Not corrupted, excel is automatically converting long numbers into scientific notation.

To fix:

Right click on ISBN column header, choose "Format cells", "Custom" and choose the option that says 0 (second option below General).

KraigBinkowski commented 3 years ago

That is odd that the bib entry has no linked object. I am hopeful that it is an anomaly. I will check the book when I am next in the library.

flapka commented 3 years ago

Right click on ISBN column header, choose "Format cells", "Custom" and choose the option that says 0 (second option below General).

Brilliant. Thanks @yulgit1 !

flapka commented 3 years ago

@KraigBinkowski As we previously discussed, I've started batch processing to match (and fill in) TMS publications data based on what's available in Voyager records. (The work is undertaken in tables outside of TMS.) A couple of questions:

Where we fill in OCLC numbers currently absent in TMS, would you like us to remove the non-numeric prefixes, such as "(OCoLC)" and "ocm"? I think they've been removed in the existing TMS data, and my sense is that this removal is advantageous.
For ISBN numbers: a. When there are two numbers, because the MARC includes both the 13-digit and 10-digit forms (e.g. "9781416576495 (hardcover);1416576495 (hardcover)"), would you like to retain both? Or just the 13-digit version? b. Do you want to retain the non-numeric text, e.g. "(hardcover)"?

KraigBinkowski commented 3 years ago

yes, please strip all prefixes
We just need one ISBN - just the 13 digit one (unless there is only a 10-digit number). No, we don't need any text such as "hardcover" if you can eliminate it. Thank you,

flapka commented 2 years ago

@yulgit1 @KraigBinkowski @edgartdata

Based on some semi-automated matching + manual look-ups by Kraig and his team, we now have a table ready to re-import back to the TMS publications module. The goal (as noted above) is to introduce a much more comprehensive and complete set of ILS and OCLC numbers. The table has three columns:

ReferenceID (TMS bib record ID): this ID can be used to match data in the table to existing entries in TMS
oclc_replaceToTMS: numbers in this column should replace data in TMS where Altnum = OCLC Number
ILSno_addToTMS: numbers in this column should be added to TMS where Altnum = ILS Number (in theory these should replace existing numbers too, but in practice few if any of the records in this set are likely to have data in ILS Number)

The full table has 3743 records to update. For the purposes of testing the process, Kraig and I wonder if it would be useful to import/overlay a smaller set, such as the 20 also attached here. Would that make sense?

Test list TMS Bib oclc ils numbers.xlsx Full list TMS Bib oclc ils numbers.xlsx

yulgit1 commented 2 years ago

This should be all set. For future reference I performed the following:

Converted Excel 97-2004 Workbook (.xls)
Used SSMS to Import into temp table AltNumUpdate052722 to run SQL cursor on
Created 2 scripts saved in 00ITdocs\000TMSdocuments\SQLqueries: RefAltNumUpdate.sql, refAltNumUpdate2.sql to filtering nulls and insert (no existing row) or update (existing row) where appropriate.

Used commands in refAltNumCheck.sql to check (143=Reference Table, 112=OCLC Number, 193=ILS Number):

SELECT * FROM AltNums where TableID = 143 and AltNumDescriptionID = 112 and LoginID like 'erj0522%' order by ID
-3104

SELECT * FROM AltNums where TableID = 143 and AltNumDescriptionID = 193 and LoginID like 'erj0522%' order by ID
-3655

edgartdata commented 2 years ago

@flapka Given the work done above, today in https://collections-test.britishart.yale.edu/catalog/tms:45 I was expecting the bib citation "Julia Marciari-Alexander, This other Eden, paintings from the Yale Center for British Art , Yale Center for British Art, New Haven, CT, 1998, pp. 16-17, fig. 16, ND1314.3 Y36 1998 (YCBA)" (aka TMS ReferenceID 45 from your list) to be linked to the YCBA Reference Library record in Blacklight via the ILS number 4226255. Maybe I understood wrong?

flapka commented 2 years ago

@edgartdata I think what you describe is the goal, right @KraigBinkowski ?

The data upload will provide the necessary ILS identifiers, but perhaps we're still working on the code that will generate the links? Is that right @yulgit1 ?

KraigBinkowski commented 2 years ago

@edgartdata The work that Eric did on the test list above just added the OCLC number and ILS number to the bib record for _This Other Eden (_TMSBib45). The linking of the bib entry to the full record is the second part of the equation after we know the fields have been added correctly.

KraigBinkowski commented 2 years ago

@edgartdata @flapka we haven't yet asked Eric to create the links or what part of the bib entry to hot link. It is my opinion that if possible, it should only be our Call Number field in the bibliography entry that links to the full record -- a bib entry in which the full entry is hotlinked seems visually distracting and perhaps confusing as online resources in the bibliography are hotlinked in this way. This is all up for discussion.

yulgit1 commented 2 years ago

@edgartdata What's done so far is only the tip of the iceberg. The ILS number would have to be indexed now. But the link would have to be the whole citation not just on the Call Number, as the Call Number is just part of the larger string that can't be parsed.

There is the callnumber field in LIDO. It would be possible to index it to solr and and concatenate it to the citation, but then the callnumber would be listed twice.

Is the full citation linked that bad? We are already doing it for online publications for example the bibliography tab: https://collections.britishart.yale.edu/catalog/tms:107

yulgit1 commented 2 years ago

One thing I noticed in Dort just sampling the first publication, see below. It's ILS number https://hdl.handle.net/10079/bibid/584227, matches the call number. But the LIDO appellationvalue is nothing like what's in the bib record. Maybe the citation is just a part of the whole?

<lido:event>
<lido:eventID lido:type="TMS">34</lido:eventID>
<lido:eventID lido:type="local" lido:source="ILS Number">584227</lido:eventID>
<lido:eventID lido:type="local" lido:source="LC Call Number">NJ18 T85 T87+ (YCBA)</lido:eventID>
<lido:eventID lido:type="local" lido:source="OCLC Number">07605713</lido:eventID>
<lido:eventID lido:type="local" lido:source="YCBA TMS Bibliographic Module record referenceID">4800</lido:eventID>
<lido:eventType>
<lido:conceptID lido:source="AAT" lido:type="local">300054686</lido:conceptID>
<lido:term>publication event</lido:term>
</lido:eventType>
<lido:eventName>
<lido:appellationValue>
^'Splashers,' Scrawlers' and 'Plasterers' : British landscape painting and the language of criticism, 1800-40^, Mallord Press, London, Summer 1990, pp. 5-11, NJ18 T85 T87+ (YCBA)

edgartdata commented 2 years ago

@KraigBinkowski https://hdl.handle.net/10079/bibid/584227 has the title "Turner Studies" in Orbis but our TMS citation with the same call number is for "'Splashers,' Scrawlers' and 'Plasterers' : British landscape painting and the language of criticism, 1800-40, Mallord Press, London, Summer 1990, pp. 5-11, NJ18 T85 T87+ (YCBA)" Could it be a mistake with the call number in TMS? Or that this citation is for an article in "Turner Studies"?

edgartdata commented 2 years ago

@yulgit1 Knowing that we will discuss this tomorrow, I wonder if this option is viable:

take out the callnumber out of the long citation in LIDO
index the callnumber separately and append it to the long LIDO citation so that it is the only hyperlink element in the citation and it is there only once

KraigBinkowski commented 2 years ago

"@KraigBinkowski https://hdl.handle.net/10079/bibid/584227 has the title "Turner Studies" in Orbis but our TMS citation with the same call number is for "'Splashers,' Scrawlers' and 'Plasterers' : British landscape painting and the language of criticism, 1800-40, Mallord Press, London, Summer 1990, pp. 5-11, NJ18 T85 T87+ (YCBA)" Could it be a mistake with the call number in TMS? Or that this citation is for an article in "Turner Studies"? "

I think this is indeed a cataloguing mistake - I will check fully and correct it.

edgartdata commented 2 years ago

There are 3 types of hyperlinks we would like to add to the bib citations:

A link from a YCBA P&D or P&S Blacklight record to a YCBA Ref Library or YCBA RBM Blacklight record
a link from YCBA P&D, P&S, Ref Library or RBM Blacklight record to a Yale Orbis/Quick Search record
A link to an external web resource, such as the Tate article https://www.tate.org.uk/research/in-focus/the-dolls-house-william-rothenstein/narrative-and-anti-narrative for https://collections.britishart.yale.edu/catalog/tms:5019

For 1. the plan is to use the ILS Number that points to a YCBA resource and then return the Blacklight URI for the resource, and hyperlink the LC Call Number with the BL URI. The LC Call Number would need to be taken out of the bib citation (we need to ask David to do that after we test our process for a couple of objects).

7838392 N5220 M552 P38 2007 OVERSIZE (YCBA) Questions: can we do that for all bib citations that have a corresponding item in either Ref Lib or RBM? Is there a risk to loose some LC Call Numbers out of the bib citation because there would not be corresponding Ref Lib or RBM resources in Blacklight? For 2. same process but point to Yale Orbis/Quick Search URIs. Questions: do we prefer Orbis? or Quick Search? @KraigBinkowski Do you have an example? For 3. I think it would be helpful if we would create a new TMS AltNum Description and lido:source type for resources available online. For this use case, we would need to ask David to take out the TMS Remarks field out of the bib citation concatenation. 5019 Available Online 3622 Question: @KraigBinkowski Are all citations marked as available online refer to non Yale resources? Feel free to poke holes in this plan!

flapka commented 2 years ago

Curious to hear @KraigBinkowski 's take on the implementation in test, e.g.: https://collections-test.britishart.yale.edu/catalog/tms:5019 -- which is in general quite promising!

I had imagined that each Bibliography entry would have no more than one link, with order of preference: YCBA catalog (if present), YUL catalog, OCLC. Having all three is interesting, in that it allows the user to specify which of the three is of interest. But how do we feel about the added verbiage? It's a little less clean to have all three.

Of more importance: it looks like we still need to implement a linking logic to account for whether YCBA has a copy of a book. In the tms:5019 example, the Timothy Hyman entry includes a YCBA link (broken) even though we don't have the item (SML does). Could the logic say if we have an ILS no and the call number includes "YCBA", then make a YCBA link; otherwise don't?

yulgit1 commented 2 years ago

"Could the logic say if we have an ILS no and the call number includes "YCBA", then make a YCBA link; otherwise don't?"

The answer is yes by indexing the callnumber as part of the ordered linked_citation_ss array.

edgartdata commented 2 years ago

I agree with Francis: I think we should only offer one link in our art object citations, which should be to our Ref Lib or RBM Blacklight record. Our RBM BL records carry the Orbis link. And we want to keep our online catalog users in our catalog as much as possible.

So I guess the next question is: do we want to offer non YCBA BL links in our online bibliography? For example, if there are no YCBA copies of the book in our Ref Lib or RBM, but there is a copy at YUL, do we want to vend this link to our citation since Kraig's team have recorded it in TMS?

edgartdata commented 2 years ago

@KraigBinkowski @flapka Also the Hyman link to the YCBA copy not working because there is no such copy, i.e. it is a film that has been classified as LC Call Number for efficiency? Kraig and I have already started talking about adding new types in TMS for such resources and it seems that it has now become a priority.

ycba-cia / blacklight-collections2

Publications list -- link entries to full bib record in blacklight. #211