Dataverse 4 (collections)

35 views
Skip to first unread message

George Duimovich

unread,
May 12, 2015, 1:00:06 PM5/12/15
to dataverse...@googlegroups.com

Just some general feedback regarding feature pulled from version 4.

In version 3.x, dataverse admins could organize their "studies" (datasets in 4.x) into collections - this functionality is missing in 4.x (at least at the presentation level - our SP dataverse not yet upgraded to 4 so don't know what's in 4.x admin).

There are search improvements for sure in 4.x, but collections was always an appealing 'customization' feature to support local dataverse browsing and always was well received in Dataverse demos I have been part of.  I see its removal as a loss of a key feature.  Also helpful for organizing diverse curated collections within a single dataverse.

Journal example:
Version 3.6.x - https://thedata.harvard.edu/dvn/dv/pan
Version 4.x - https://dataverse.harvard.edu/dataverse/pan

IMHO, the version 3 presents much more as one would expect a Journal (exposing the local collection facets), whereas the 4.x version at first glance appears more like a generic repository (not as clear that it is a journal).

Another one to compare:

Murray Research Archive  https://thedata.harvard.edu/dvn/dv/mra (in 3.6x)  vs. the 4.x version here:https://dataverse.harvard.edu/dataverse/mra

So just a quick recommendation for bringing that functionality back to DV 4.x, especially helpful imho for adding custom local browse to Dataverses.

Any technical or other reasons for the withdrawal of this feature 4.x?

George Duimovich
MacOdrum Library
Carleton University




Philip Durbin

unread,
May 12, 2015, 2:48:30 PM5/12/15
to dataverse...@googlegroups.com
Hi George,

Thank you very much for the feedback. I'm hoping others on the team can jump in but I think the "Migrate Collections For Dataverses Using New Dataverse Functionality" issue is where some early thinking/development took place: https://github.com/IQSS/dataverse/issues/121 . (Journals seem to be mentioned in particular.)

I see both the Murray Research Archive and journals mentioned in this issue which is still in development: https://github.com/IQSS/dataverse/issues/1549

I hope this helps a bit at least...

Phil


--
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 post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/60449339-b0d4-42fe-b43f-88162926d830%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

Mercè Crosas

unread,
May 12, 2015, 4:20:41 PM5/12/15
to dataverse...@googlegroups.com
Thanks for the feedback. We have some ideas to show the hierarchy of sub-dataverses within a dataverse, which we'll track here: https://github.com/IQSS/dataverse/issues/2155

The MRA dataverse and journals dataverses need some post 4.0 curation work to be better organized. This work is in process and will be improved soon.

Collections didn't work well across dataverses. Once all the features are in place in the next 4.0.x releases, the sub-dataverses and facet functionality will be a nice replacement for the old collections, at least that's our vision, but we'll test it and keep improving it as needed. One of the problems with Collections is that they were not part of the structure or  metadadata of datasets, so they did not live with the dataset record (which is not good for discoverability, and interoperability with other repositories). The solution with Dataverse 4 tries to solve this problem.

Merce


Mercè Crosas, Ph.D.
Director of Data Science, IQSS
Harvard University

Reply all
Reply to author
Forward
0 new messages