Publication citation linking + Make Data Count questions

[Amber Leahey]

Hi everyone, 

I'm hoping these are questions others have dealt with. Any help clarifying is appreciated! 

1. We haven't implemented Make Data Count yet but we are testing it in our demo instance. If we set this up in production will it wipe out our current metrics? Is there a way to configure Dataverse to use both types of metrics? I don't see this in the Harvard Dataverse but I could be missing something. 

2. If a depositor adds a related publication will this automatically be linked to the article via DataCite/CrossRef (not exactly sure how this is working and where these dataset citations will appear from publications, on publisher websites, Web of Science?) OR do we need Make Data Count for this to happen?

Many thanks,

Amber Leahey

[danny...@g.harvard.edu]

Hi Amber, I'll try to answer these below. 

1. This is a challenging question :) The display will only show the MDC counts. Our plan was to be able to migrate legacy counts in such a way that only one number would be displayed, but we haven't yet been able to achieve this in Harvard Dataverse Repository because we haven't been able to prioritize the migration. I believe QDR has display for both types of metrics (see https://github.com/IQSS/dataverse/pull/6543#issuecomment-638246897) but this is not something we've brought back into the core codebase because we'd rather solve this with a single number. It's just been a matter of prioritization and we have not yet turned on MDC in Harvard Dataverse Repository because of this. We would like to come back to the MDC work in the future, which would include providing some better reporting and some better tools to get the legacy counts moved over.

2. I believe MDC is currently the only way to do this. There's some discussion about providing better support for this here: https://github.com/IQSS/dataverse/issues/2778 but it hasn't been prioritized. 

- Danny

[Philip Durbin]

To follow up on Danny's points, here's the issue we're using to migrate legacy counts into Harvard Dataverse: https://github.com/IQSS/dataverse.harvard.edu/issues/75

Like Danny said, QDR has implemented showing both classic and Make Data Count download counts in their fork. Here's a live example with a fair number of views and downloads: https://doi.org/10.5064/F68G8HMM

For the question of "If a depositor adds a related publication will this automatically be linked to the article via DataCite/CrossRef?" unless I'm turned around, I don't think Make Data Count will help here. Make Data Count is about tracking views and downloads. It's also about retrieving dataset citations from DataCite/Crossref. As Danny points out, https://github.com/IQSS/dataverse/issues/2778 has the most discussion about pushing the "Related Publication" relationship to DataCite. It's a little hard to follow giving the number of comments but I think it's up to Dataverse to send the right metadata to DataCite when a dataset has a relationship to a related publication.

I hope this helps,

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 view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/9d0bbbc7-d27c-4aba-9d3c-9f4b0c805e2an%40googlegroups.com.

--

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

[James Myers]

A little more info:

1) The challenge in combining the numbers is that MDC, as implemented in Dataverse with counter_processor, does a good job of removing requests from robots, separates machine and human accesses, only counts rapid repeat requests from the same user as one count, etc. The internal accounting in Dataverse prior to MDC didn’t do such filtering and the base guestbookresponses table (where non-MDC counts come from now) doesn’t really have enough info to do such filtering. If you have other logs (such as the Apache log or optional glassfish/payara logs) you might be able to use those, possibly in combination with the guestbookresponses to do some filtering. Harvard has been looking into that. For QDR, we did not have complete apache logs and didn’t want to just combine the unfiltered pre-MDC counts with the new MDC ones, so, in cases of datasets with counts prior to when MDC logging was turned on, we made a code change to show both. (If you recall, I was pushing people to turn on MDC logging even if they weren’t ready to turn on display of MDC and to set up all the daily/weekly processing. This was the reason – it’s only the downloads prior to the start of MDC logging that would disappear in the community Dataverse releases, and only things prior to the MDC logging start would have to be custom processed (ala Harvard) or handled by a code change (ala QDR)).

 

2) is ~independent of MDC – right now Dataverse isn’t reporting related things entered as Dataverse metadata to DataCite. QDR has  added that, making some choices about which fields mapping to which relationships.( I think #2778 is awaiting consensus on such a mapping and then whatever minor coding is needed to enact it). If/when Dataverse sends that info to DataCite, it could, in theory be picked up so that the paper could indicate a connection back to the data in Dataverse (I’m not sure what’s actually in place at CrossRef/elsewhere w.r.t. this). MDC is actually involved in the opposite direction – if you enable MDC and run the weekly script, Dataverse will pick up any instances where a paper cites data in Dataverse and display those citations in a list on the page.

 

-- Jim

--

[Gautier, Julian]

About related publication data citations, OASPA is hosting a webinar on July 13 about Make Data Count. The webinar seems to be focused on journal publishers and about data citation specifically. Anyone can attend. I plan to listen in - thinking it'll be helpful to learn how journal publishers are and are being encouraged to send data citation info (like the use of those relationship types Jim mentioned). More info about the webinar and registration info:

---------- Forwarded message ---------
From: OASPA <in...@oaspa.org>
Date: Wed, Jul 7, 2021 at 5:14 AM
Subject: Reminder for next week's OASPA Make Data Count webinar

View this email in your browser A reminder that our interactive webinar hosted in collaboration with Make Data Count is happening next week (July 13). The webinar will be centered on best practices for data citation. We want to understand and help remove barriers to data citation, and support those already doing this valuable work. As leaders in the evolving open access publishing space, OASPA believes that developing and supporting best practices and broadly achievable standards for all aspects of open content is essential, including for research data. Because of this, OASPA endorses the Make Data Count initiative, and its goals of building for and supporting the inclusion of research data in responsible research assessment.  Please see the blog post to learn more and join us for the webinar to contribute to this important work. Best wishes, Bernie Bernie Folan Communications, Engagement and Outreach Manager, OASPA bernie...@oaspa.org Copyright © 2021 Open Access Scholarly Publishers Association, OASPA, All rights reserved. You subscribed to our newsletter from our website. Our mailing address is: Open Access Scholarly Publishers Association, OASPA Prins Willem-Alexanderhof 5 The Hague, 2595BE Netherlands Add us to your address book Want to change how you receive these emails? You can update your preferences or unsubscribe from this list.

--

Julian Gautier

Product Research Specialist, IQSS

ᐧ

ᐧ

ᐧ

[Amber Leahey]

Thank you for your replies! 

Thanks for the encouragement to turn on MDC even without it being displayed yet. 

Regarding sending linkages to DataCite, we have our first Journal collection and they really would like to see this. 

I'll follow up on some of those github tickets to see what the latest thinking is on moving this along. 

Best, 

Amber

[Julian Gautier]

DataCite is hosting a series of webinars about Make Data Count, and like the webinar that OASPA hosted in July, I'm hoping these webinars are an opportunity to make data driven decisions about how we and other publishers use  relation types to link related resources. The link to register for the first webinar on March 17 is at https://datacite.zoom.us/webinar/register/5516442572756/WN_R-3E4OlwTJ6p97RDi_-Wog. The times and titles of the webinars are in the image below:

[Julian Gautier]

For anyone who's interested but missed the first webinar on March 17, yesterday the DataCite folks posted a video of it at https://www.youtube.com/watch?v=dy1Z4vGXHxo.

Registration for the second webinar is at https://datacite.zoom.us/webinar/register/1216481206599/WN_Vs3UsbaXRG6zPxEIjC89YQ