Current version of the Dataverse Metadata Crosswalk

[Philipp Conzett]

I've been asked  by one of our partner institutions whether the Dataverse 4+ Metadata Crosswalk is the most updated one. I couldn't tell. Can someone help? Thanks!

Best, Philipp

 

[Julian Gautier]

Hey Phillip. Yeah, that's the most up to date one. There's no other crosswalk we maintain and as far as I know the mappings to fields in various exports that Dataverse creates by default are up to date, including all of those question marks :o

I plan to update the crosswalk to reflect changes in an upcoming Dataverse release, but those will be changes to some field labels and not changes to Dataverse's default mappings.

Julian Gautier (he/him)

Product Research Specialist, IQSS

Interested in helping test Dataverse? Sign up for usability testing

[Julian Gautier]

I hope knowing that helps :)

There's a note in the first "About" tab about how the crosswalk represents the most up to date mappings. But I know it can be easy to miss tabs in spreadsheets, especially if the link someone uses goes directly to the second tab. And maybe the note in that "About" tab doesn't help folks feel certain that the crosswalk is the most up to date?

It would be easy to add a last updated note somewhere, maybe as part of the spreadsheet's title, such as "Dataverse 4+ Metadata Crosswalk (last updated 2023-##-##), so it's easier to see. We could even remove the "4+" in case folks see that and wonder if it means that it reflects mappings in Dataverse 5 and 6 and so on.

What I think would take more discussion than I've had time to encourage is how else to ensure that the crosswalk gets updated and remains otherwise remains helpful. I've floated the idea of versioning it in sync with Dataverse releases, and I wonder if we could somehow require that when someone makes changes to the default mappings, they also update the crosswalk. Sometimes I'm not aware of changes until much later.

[Philipp Conzett]

Thanks for the heads-up, Julian! Once operational, maybe the EOSC Metadata Schema and Crosswalk Registry (MSCR) will be a good option to define the Dataverse metadata crosswalk: https://faircore4eosc.eu/eosc-core-components/metadata-schema-and-crosswalk-registry-mscr.

Best, Philipp

[Philip Durbin]

Crazy idea I may have mentioned before: add the ability to query Dataverse for this crosswalk information via API.

That is, you ask Dataverse for a crosswalk for the "title" field and you get back:

DDI Codebook 2.5: stdyDscr > citation  > titlStmt > titl
DC Terms: dcterms:title
Schema.org JSON-LD: name
etc.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/4de64b77-7dc5-4391-b678-49110ffbea9an%40googlegroups.com.

--

Philip Durbin
Software Developer for http://dataverse.org
http://www.iq.harvard.edu/people/philip-durbin

[Philipp Conzett]

@Phil: Do you mean querying individual Dataverse installations?

[Philip Durbin]

Yes, exactly. Just a crazy idea. :)

To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/0a769d22-870c-44e8-9880-02d2fb32e7a6n%40googlegroups.com.

[Philipp Conzett]

Sorry for late reply. Wouldn't the mapping be the same in all installations?

[Philip Durbin]

Well, sure, all installations running the same version of Dataverse with the same metadata blocks loaded up would have the same output.

To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/65e934d8-9666-4f3b-82b3-b5c882b83db0n%40googlegroups.com.

[Julian Gautier]

Managers of installations running Dataverse v5.14 and later can change how fields are mapped in the metadata exports that ship with Dataverse and create new exports, though I'm not sure how many installations are using this feature. So installations might not have the same output.

I think Phil's crazy idea of being able to use the API to ask an installation how each field is mapped to fields in certain exports would be helpful in creating a crosswalk. It could automate a lot of what I've been doing so far, which is creating a dataset with all fields filled, then looking at the exports and at how the metadata is imported in different ways (such as SWORD, API, OAI-PMH). And it might help us see some mappings that I haven't been able to determine.

There are notes in the current crosswalk that would have to be recorded somehow, like how a Dataverse field is mapped to different fields in the same standard, like Dublin Core, depending on how the metadata is imported or exported. But overall I think Phil's crazy idea would be pretty helpful.

[Sherry Lake]

Hi Julian,

I did not know about creating new exports. Could I use this feature to create new exports based on existing ones. For example, if I wanted a more "complete" export of DataCite metadata, could I create a new export "DataCiteExpanded" (?) that includes additional metadata fields than the default?

I'm thinking about ways to enhance what we send to DataCite, such as Related works, so the "links" between works and datasets can be more connected.

Does this make sense?

Thanks,

Sherry Lake

To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/f114acdb-bda1-40ea-9f61-1ddc3284e5b0n%40googlegroups.com.

[Julian Gautier]

Hi Sherry,

Yeah you could use the feature to create new exports based on existing ones. But there might be other steps, outside of the feature, for changing the metadata that's exported to DataCite when registering DOIs and publishing new versions.

[Sherry Lake]

Hi Julian,

Yes, I know how to update DataCite DOI outside the Dataverse software, I'll have to use APIs to do the DOI updating. I can do that; I have done that for other DOI registrations from our other repositories.

It would be easier if I could start with a json file (exported from Dataverse) to use in my API call.

Thanks for bringing up the outside of the feature "extra" part.

--

Sherry 

To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/48ddf9ec-fb92-4d0d-a19f-a8349f6374d6n%40googlegroups.com.

[Julian Gautier]

I like the idea of being able to use the API to query Dataverse for this crosswalk information. For example, I could imagine writing a script that uses such an API to update the crosswalk at https://docs.google.com/spreadsheets/d/10Luzti7svVTVKTA-px27oq3RxCUM-QbiTkm8iMd5C54, which represents how "out of the box" Dataverse imports and exports metadata. Then I could run the script every time a new version of Dataverse is released, instead of updating that crosswalk "by hand". The GitHub issue at https://github.com/IQSS/dataverse/issues/4451 is kind of related.

And the API could help managers of Dataverse installations that have different metadata exports create and update their installation's crosswalks, too.

[Vyacheslav Tikhonov]

Hi Julian and others, some time ago I have created graph representation for metadata schemes, can it be useful?

https://github.com/Dans-labs/semaf-client/tree/master/schema

Best,

Slava

[Julian Gautier]

As someone who maintains a crosswalk, I don't know how I'd use the files I see at https://github.com/Dans-labs/semaf-client/tree/master/schema for the tasks that I hope any crosswalk would help me with. But that's probably because I'm not familiar with how those files are created and how they could be used.

It's very timely that you've broadened the scope of this conversations after Phillip shared that Survey on Semantic Artefact Mappings and Crosswalks in that other Google Groups thread. Just based on my own experiences, I'd expect the results of that survey to show that most folks use spreadsheets to create and maintain this information just because most people are familiar with rows and columns. There's a low learning curve and ways to mitigate at least some of the awkwardness of using them, like needing to flatten important info into just two dimensions.

I'd second Phillip's call to contribute and learn from that survey about why and how others create and use tools like crosswalks.