Proposal: Improving dataset relationships in Dataverse

[Vera Clemens]

Hello everyone,

we would like to announce that we intend to work on a new feature – improving structured dataset relationships in Dataverse. This feature aims to enhance the way related datasets are managed and displayed.

You can find the detailed proposal here: https://docs.google.com/document/d/1VVF2v8OGB1LCN5XLG93tK6Lbz2DuP5AmefOQhbWtkEQ/edit?usp=sharing

If you're interested in collaborating or providing feedback, please reach out or join us during the upcoming BioHackathon Europe (Nov 4 - 8, 2024, https://biohackathon-europe.org/), where we will be discussing and working on this feature.

Best regards,
Vera

(Software developer @ NFDI4Health, https://www.nfdi4health.de/)

[Philipp Conzett]

Hi Vera & team,

Thanks for initiating this work and announcing it here! This sounds like a useful feature to enhance linked data support in Dataverse.

Looking forward to hearing more about the project!

Best,

Philipp

[Julian Gautier]

Hi Vera,

I'm very interested in collaborating and plan on joining the call on Nov 4.

It's possible that more work will be done before that Nov 4 call. The Dataverse UX working group plans to tackle this, too, and the Dataverse team has been working on this topic with other repository products through a multi-year, NIH-funded initiative called GREI. So as you've implied in this post, I'd like to stress the importance of collaboration and outcome-driven development.

All best,

Julian

Julian Gautier (he/him)

Product Research Specialist, IQSS

Interested in helping test Dataverse? Sign up for usability testing

[Julian Schneider]

Hi Julian,

Sounds great! Vera and I will attend the BioHackathon in person - do you intend to join us there, or remotely?

We would gladly speak to you to align our ideas about this topic beforehand. Maybe we can simply join you in the regular UX working group meeting, or otherwise schedule a separate meeting for this?

We're unavailable next week, but could make time starting Sep 23rd.

Cheers,

Julian

[Julian Gautier]

Ah, I don't know why I thought there was a remote call on Nov. 4. Sorry for the confusion. I see that Vera mentioned the BioHackathon Europe conference on Nov 4 - 8, 2024 and I definitely couldn't attend in person. I can try to attend remotely, depending on the timing and how I might contribute during a hackathon.

I think it would be great to discuss before then since the proposal you shared includes a timeline with an analyze and design phase happening this month and in October.

We couldn't meet about this during the upcoming regular UX working group meetings, since we're in the weeds on the "sketch" phase of a redesign of other parts of the citation metadata block.

Could we find a time to meet in a separate remote call as early as next week, like you wrote, or the week of September 30? Unless you'd like to, I could send a poll around to see what times work best. Let me know!

In the meantime, I thought it would he helpful if I shared the scope and approach of the work we're planning. It's in a proposal at https://docs.google.com/document/d/1uXk0KTfnZuRH2Iw3OTZEzP6YgV5O6fX2JPRQdWZBiTY, particular pages 3-4. Not as detailed because there are parts of the approach that will include discovery research (and because we've been very focused on the first part of the proposal about improving metadata about people and organizations!)

Thanks!

[Julian Gautier]

Hi again. I thought I'd share an article that's related to this topic, for anyone who hasn't seen it yet, and that I've been slowing making my way through. The article's at https://www.arxiv.org/abs/2408.14636 and co-authored by Natasha Noy at Google. Years ago Natasha helped us a bit with the design of Dataverse's Schema.org export. So far the article seems like a helpful overview of the reasons for, challenges with, and approaches to identifying relationships between datasets.

[Clemens, Vera]

Hi Julian,

 

we’ve gone ahead and created a poll to hopefully find a suitable time for a meeting either next week or the week after that: https://www.when2meet.com/?26722043-knkgA

 

Thanks for sharing the link to your proposal, as well as the related research article, looks very interesting!

 

Thanks and best regards,

 

Vera

--
You received this message because you are subscribed to a topic in the Google Groups "Dataverse Users Community" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/dataverse-community/uX-uLQ5EEXM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to dataverse-commu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/821b0189-4084-450c-9182-43a7f2bf4a0en%40googlegroups.com.

[Julian Gautier]

Awesome, thanks Vera! I filled out the poll.

[Vera Clemens]

Hi,

as planned, a prototype has now been created :) I have posted some screenshots of the prototype in the related GitHub issue: https://github.com/IQSS/dataverse/issues/5277#issuecomment-2462342497

Happy to receive feedback on the current state!

I'd like to do some more cleanup, after that I will also share the prototype code.

Best regards, Vera

[Vera Clemens]

Hi,

the prototype code is now available here, along with installation instructions and some updated screenshots: https://github.com/vera/related-datasets-cvoc

Best regards, Vera

[Philip Durbin]

I finally got around to playing around with the prototype late last week. I like how related datasets can be local or any arbitrary remote URL.

Also, Vera and I are talking in Zulip if anyone would like to join us there: #dev > Improved "Related datasets" @ 💬

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 visit https://groups.google.com/d/msgid/dataverse-community/48a3c112-1723-4082-8e61-54358e4f5e10n%40googlegroups.com.

--

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