Change source for the ISBN-Search by Identifier

Juergen

For the search with certain unique identifiers Zotero uses WorldCat as source for ISBNs. Is it possible th change the source and to add another database, for example the german catalogue "GVK - Gemeinsamer Verbundkatalog" (http://gso.gbv.de/)?

Thanks for response

Juergen

kselmiwlb

5 years later, I am asking myself the same question...
If I research in my library's catalogue and import the metadata, they are retrieved not from my library but from WorldCat. Is it possible to change that?
Thanks for your response
Katja

adamsmith

Katja - how are you importing your metadata? By inputting the ISBN or using the URL bar icon? What's the catalog you're using?
(As for the original question: no, that's not possible. We do, though, currently check the library of congress before we go to WorldCat. We're considering adding GBV, but don't want to add too many layers).

kselmiwlb

Hello and thanks for your quick response.
I import my metadata from my library catalogue https://wlb.ibs-bw.de/ by clicking on the URL bar icon.
Then in my zotero record, in category "library catalogue" it says "Open WorldCat" or "Library of Congress ISBN".
Now I've just noticed that when I import from other German libary catalogues, zotero does seem to import the metadata from the catalogue itself; then it says e.g. "Library Catalog - iaiweb1.iai.spk-berlin.de" or "Library Catalog - stabikat.de".
Could that be a problem of our library catalogue? Maybe it doesn't communicate well with zotero, and so zotero uses the ISBN and searches in the LoC catalogue and WorldCat instead?
Katja

zuphilip

Corresponding ticket: https://github.com/zotero/translators/issues/737

adamsmith

Maybe it doesn't communicate well with zotero, and so zotero uses the ISBN and searches in the LoC catalogue and WorldCat instead?

yes, this. When using COinS, if there is an identifier (DOI or ISBN) Zotero uses that to fetch metadata for the item in question. In general that's a good idea, since that data is better than whatever could be embedded in COinS.
As per the ticket zuphilip links to, it would certainly be possible to improve this for the aDis catalog family, but it's a fair amount of work and no one has gotten around to it yet.

kselmiwlb

Ok. In general, the data from WorldCat and LoC are ok. But in some cases the WLB catalogue metadata are better, and then it's confusing for users that when they import the title into zotero, they do not get the same data they see in the corresponding library catalogue record.
A translator for aDIS catalogues would be nice. There are quite a few libraries that use aDIS, especially in the southwest of Germany. But I know that would mean a lot of work...
Thanks again for the information.

adamsmith

It's your state, zuphilip ;)

zuphilip

@ksemliwlb : You are a librarian at WLB in Stuttgart, right? Can you change the behaviour of your catalogue or is it hosted by the BSZ?

What actually happens at the moment is the following: The zotero picker (url bar icon) is using the COinS data from your catalogue. This is incomplete, as the authors/contributors are always missing. Then the COinS translator will look the same item up in several other catalogues (e.g. LoC, WorldCat) to receive better metadata.

Thus, we should first try to improve the COinS data in your catalogue. The advantage is that also the Citavi picker can read COinS data and therefore could be helpful for Citavi users as well. Moreover, your RIS exports seem to begin with some strange symbols ïůż and bibtex seems not working reasonable:

@book {
author = {Breindl, Eva and Volodina, Anna and Waßner, Ulrich Hermann},
title = {Semantik der deutschen Satzverknüpfer},
address = {Berlin},
year = {2014},
publisher = {De Gruyter Mouton},
pages = {XII, 681 },
URL = {http://deposit.d-nb.de/cgi-bin/dokserv?id=4561711&prov=M&dok_var=1&dok_ext=htm},
ISBN = {978-3-11-034134-8},
ISBN = {978-3-11-039453-5 (Sekundärausgabe)},
}
U1 - Hauptbibliothek
ER -
%@ 978-3-11-039453-5 (Sekundärausgabe)
%[ 20140826
%= 20141128
%1 Hauptbibliothek
Here are some background information how to expose metadata such that reference management tools can grab them. @kselmiwlb: Let us know if you can change something from your side and feel free to write me also at my work email adress. We also understand German if that is easier for you at any point.

kselmiwlb

Hi zuphilip, yes, I am a librarian at WLB.
Testing a new version of our catalogue, we have noticed in the meantime that the import works just fine there. I suppose that the coins data have been improved for this version (which is not online yet, it will be so within a few weeks.)
We know about the problems with our RIS export. If you remove the "strange symbols" at the beginning of a RIS export text, you can import the data... Our IT is in contact with the catalogue developers in order to solve this problem.

adamsmith

that sounds promising - let us know when the new catalog is up and we'll re-evaluate whether writing something specific for it makes sense or whether the COinS is good enough.

zurpher

Hello,

it seems that I have imported GBV data when using " Add Item(s) by Identifier " option. Great! I particularly like that the ToC is imported as well.

Three things that could do with some improvement:

Only the first editor of five has been imported. Adding the other four is a bit cumbersome.

The language field has " ger " as a value.
Could that be change to " de "?

When there is more than one place for a pulisher, it is abbreviated with [u.a.]. This is a German abbreviation and I always have to delete it because I write English papers as well. Furthermore, the citation styles that I use only require the first place.

zuphilip

Yes, you are right that the GBV search translator is in place now.

1) Can you give an example with five editors?
2) The codes should come from this list: http://www.loc.gov/marc/languages/language_code.html . Thus, I guess we could map at least the common languages spoken today. @adamsmith, @aurimasv: Do you agree? Is the Marc translator the correct place to do this?
3) That is a tough one and I am not sure how much we can/should trying to change during import of metadata.

adamsmith

2) I'd be inclined to move the mapping function for languages out of the translator into a utility. We generally don't fix the language in translators, though. The idea so far has always been that we'd do this in the database when we actually use the language field for anything other than turning off title casing (which ger and de will do equally).

zurpher

1)

Böttcher, W., Kerlen, C., Maats, P., Schwab, O. & Sheikh, S. (Hrsg.). (2014). Evaluation in Deutschland und Österreich. Stand und Entwicklungsperspektiven in den Arbeitsfeldern der DeGEval - Gesellschaft für Evaluation. Münster: Waxmann.

http://gso.gbv.de/DB=2.1/PPNSET?PPN=792088727

2)
@adamsmith: You're right. For my purposes, it actually doesn't make a difference.

3)
I wish it would be "fixed" on GBV's side but perhaps that's too much to ask. I wonder whether there is any scenario where having the [u.a.] makes sense.

zuphilip

1) The other editors are not in the data from the GBV. This seems to be the old-fashioned cataloguing rule, for more than three creators/contributors. Thus, we cannot improve that in Zotero. (However it might improve on their side in the future with RDA coming next year.)

3) I could delete "[u.a.]" at the end of the publication place in the GBV translator (I guess that is better than in the MARC translator). However, if more places are given then they also should show up in zotero. @adamsmith: Do you agree on this changes?

adamsmith

3) happy to have the [u.a.] removed -- are there rules for square brackets in MARC records? Could we delete everything in square brackets in the place field (and do that in the MARC item) or will that lose us data?

zuphilip

I wouldn't do that. Normally, square brackets are used in cataloguing to indicate that the information was derived from somewhere else and not from the book lying in front of the cataloguer. This might also occur in the publication place field for information we want to keep.

adamsmith

just removing [u.a.] in GBV is fine then.

zurpher

(However it might improve on their side in the future with RDA coming next year.)

Might? Is there a way of bringing this to someone's attention? Do you know who that someone might be?

http://www.gbv.de/kontakt

zuphilip

Well, RDA will come next year in the whole German speaking area (D+A+CH) and RDA is currently already implemented in US, UK, Canada, ... It will change several things and among them is AFAIK the handling for multiple authors (maybe also editors).

I don't think it is needed to bring this issue to someone's attention. Actually, I rather expect they have discussed this already, but I don't know details about it. Maybe, we can just wait and try to see how it will be handled.

zurpher

Okay. At my library they even don't provide COinS data up for grabs. When pointing it out to them, they write that they currently have more important things to do. Mysterious.

zuphilip

Looking at the example on page 7 of this presentation it should change. However, I didn't read the remarks there carefully and I am not sure if this are the international rules or just some national rules.

zurpher

I guess it's going to be a lot of work to add authors and editors retrospectively. It's certainly a positive change though. :)

zuphilip

New rules are normally applied to new titles only and maybe to old titles when modified.

DWL-SDCA

@zuphilip While I'm not a cataloger, I try to actively follow cataloging issues as they are relevant to data standards.

There are utilities within several library automation software packages that 1) will help with the crosswalk from one cataloging standard to another; and 2) commercial and crowdsourced databases that can provide new record versions of old titles. I have no idea about the quality or completeness of these projects.

I think I remember reading that the OCLC is working on this.