Death by a thousand cuts - Wikicite and Wikidata

[Gerard Meijssen]

Hoi,

At some time Wikicite was alive and well. Now people at Wikidata state that given that the roadmap of Wikicite has not been updated for a long time, it is presumed dead. [1] As a consequence it is all too easy to ask for the "cleanup" of the existing scholarly data and imho mis-representing what has gone before.

In the years since the last WikiCite roadmap update, a lot has changed. 

• Magnus rewrote many of his tools in RUST, including the SourceMD tooks, it made no difference for the community
• Elsevier has opened up its references; they are now available at Crossref.
• Scholia now knows where a paper is used as a reference in particularly the English Wikipedia
• Scholia templates exist on many subjects and scientists in the English Wikipedia
• Wikidata is now used to improve the information of the papers used as references with information from Wikidata
• There was an initial run linking books know by their ISBN from Wikidata to Open Library. 

Personally I still add papers, one at a time, and use them as "cites work" references. For books I add the books and often link to Open Library.. I care about ecology, rewilding and when I feel compelled to work on a specific paper, I will. [2] When I come across a scientist who is in the news, I will use the author-disambiguator to link to its papers.

The last I heard about plans for Wikicite was what to do next centred around the notion that we "could" have all the papers in a Wikibase. As far as I am aware, whatever happened is not generally known and it may be a lot but I expect nothing much; I prefer to be surprised.

When these people who have their own pet projects get their way, it will destroy all the work that has been done. It will destroy mine. The notion that it will be for the better can be understood from their perspective. My problem is that it will make Wikidata only more biased. When you compare any subject that has a worldwide validity, its coverage is dominated by what we know and it is North American, European. You are unlikely to find any city of Africa with all its mayors. We do not know all the national ministers for the twenty first century and obviously not for the twentieth century of the African countries. 

For Wikicite is to be alive and well it needs to have a goal. For me it is for the all the references to scientific papers to be known in Wikidata, including the papers they cite, including its authors. This will provide a rabbit hole where people can find additional material on a subject. In addition it will show when the science referenced in a Wikipedia article is out of date. It happens and old ideas are jealously protected.

So what will it be.. Is there live in Wikicite?

Thanks,

       GerardM

[1] https://www.wikidata.org/wiki/Wikidata:Project_chat#Cleanup/Import_of_scientific_articles_and_our_notability_criteria

[2] https://scholia.toolforge.org/work/Q105451449

[Roderic D. M. Page]

Hi Gerard,

I'm not aware that people think WikiCite is dead!? 

My own ideal is to have every taxonomic paper in Wikidata, focussing on papers that are the first description of new species. Related to this is the goal of (a) having every taxonomist in Wikidata, linked to the papers they published, and (b) all taxonomic publications linked to the names they publish.

This is obviously a small subset of the goal of "all scientific papers" (and why just "scientific", that excludes a lot of academic work).

It's not clear to me in what way "pet" projects will destroy previous work (apart from the suggestion that we should delete authors known only as strings not things, which strikes me as ill thought out vandalism). I think that better coverage will come in part from "pet" projects. For example, the taxonomic literature spans centuries and multiple languages. I can't read or write Chinese, but I'm adding a lot of Chinese biology papers because they include species descriptions. Hence the Chinese-language content increases, even if that is not the focus of my project.

Anyway, I'm curious as to whether others share your concerns about the future of Wikicite.

Regards,

Rod

[Liam Wyatt]

Dear Gerard, 

Firstly, let me state that I do not think it is appropriate that many individuals' work should be described so dismissively as you have done in your message. The ability to bring together the good-faith efforts of many people toiling-away (often by themselves as volunteers) in the many disparate corners of the wikiverse to make a coherent whole, is a valuable *feature* of our movement. Please do not denigrate the valid work of others as mere 'pet projects' that will 'destroy' your work. Raising UP some area of work for attention is good, but please don't do that by pushing DOWN others.

With regards to the wider topic of *is WikiCite active* or still alive:

This depends a lot on your definition of what 'WikiCite' is...

- For some people WikiCite means creating WD items about scholarly journal articles and (arguably more importantly) the inter-connecting of these items to each-other to other WD items. Lots of people have done lots of interesting work in this field, and it continues. We are all aware of the debates about what the technical and project-scope limits of Wikidata about this topic - and these are good debates to have, in order to keep our eyes 'sharp' to the needs, risks, and possibilities of our projects. 

Either way though, this is a thing which is active.

- For some people WikiCite is the community of people who are working on those kinds of issues - the people in the WikiCite mailing list, on talkpages, and various social media groups, and in the associated WikiCite telegram group: https://t.me/joinchat/HEWPXpqR0U74yDtK
While the WikiCite mailing list is low-volume, the telegram group is quite active. (as are various other fora which are not specifically designated for WikiCite-type discussions but are nonetheless related).

- For some people, WikiCite means a conference series of that name to discuss citations in Wikimedia. There were three physical editions (2018 - Berkeley, 2017 - Vienna, 2016 - Berlin). On this October just passed we held a virtual conference in parallel to the WD Birthday events. https://meta.wikimedia.org/wiki/WikiCite/2020_Virtual_conference This had 32 hours of presentations delivered by 82 speakers, over 3 days, in 14 sessions held in all timezones, and hosted in 5 languages (with individual presentations from a further 2 languages), broadcast onto 9 different social media channels live, and now also archived on Commons. 
I would consider that to be active. 

- For some people, WikiCite is the offering of community grants for work relating to citations in Wikimedia (broadly defined). In the current round, there are 23 grants funded under the heading of WikiCite (some of which are already concluded). These are described here: https://diff.wikimedia.org/2020/10/22/wikicite-awards-23-grants-escholarships-to-improve-open-citations/ They feature a variety of focus languages/countries and wikiprojects (not just Wikidata). They are diverse in their purpose: some are software development, some are about documentation of workflows, some are about translation/localisation of tools, some are about cataloguing and digitisation of rare source materials, some are about training events, some are content creation in areas with low Wikimedia coverage.
I am also proud that this project allowed the creation of the wikiverse's first living-allowance grants: "eScholarships" to stay at home during covid and be financially supported for a few days to do valuable wikiwork, with a value defined by living-location calculation rather than purchases, and no receipts required. I feel this is an innovation in Wikimedia grantmaking that WikiCite developed which speaks directly to the idea of equity in funds dissemination, and therefore the strategic goal of knowledge equity.

I consider this to be very active.

- For some people, WikiCite is the name of a 3 year grant - from the Sloan Foundation – to fund the previous two things I've described (and funds me part-time, previously Dario Taraborelli, to coordinate them). This grant is coming to its conclusion in the middle of this calendar year. I will be creating the final report for the Sloan Foundation about that work and publicising it in due course, and it will eventually live alongside the previous reports: https://meta.wikimedia.org/wiki/WikiCite/administration#Annual_reports Originally this grant was to focus primarily on in-person events. Naturally, we had to change that focus in 2020, which required radical rethinking on our side, and flexibility on Sloan's side. To conclude this grant, the steering committee and I are currently scoping for commissioning a research report into the "state of citations" across Wikimedia, which we hope will provide a valuable snapshot into the state of our movement's use of citations at this time.

So, in that, I consider this to be active. Although, of course, once the current grant concludes this aspect of WikiCite will formally conclude too. 

- For some people, WikiCite meant a roadmap towards creating a 'bibliographic corpus' of citations - with various options for how (and where, and at what scale) to do that. This was most 'formally' stated in the form of several options devised in 2018. https://www.wikidata.org/wiki/Wikidata:WikiCite/Roadmap The existence of this document is sometimes mistaken to imply that specific staff and hardware resources had been allocated, by someone, to build "it". That is not the case - the WikiCite grant (described in the previous point) was to support events and outreach, not software development. I am currently trying to develop a future roadmap - which I won't go into in this already long email. Nonetheless, as you've stated in your original message, many things have changed in the wikiverse and in the wider bibliographic industry since 2018 which have helped increase the amount and the interconnectedness of citations (on Wikidata and elsewhere). The grant named "WikiCite" can't/doesn't take credit for those changes; nonetheless, there is momentum in the academic world for Open Citations - which is certainly positive! 

And in that final sense, the goal of WikiCite to support Open Citations is very active :-)

Liam Wyatt [Wittylama] WikiCite Program Manager Wikimedia Foundation

--
Meta: https://meta.wikimedia.org/wiki/WikiCite
Twitter: https://twitter.com/wikicite

[Gerard Meijssen]

Hoi,

I totally agree. But you mistake the messenger for the message. It is not my message; I bring it. You can verify my credentials by looking at my contributions. 

The problem with both Wikidata and Wikicite hiding on Telegram is that for reasons they deem to be significant we have lost the conversation and lost their conversation with others. There is also a Facebook environment where people talk about Wikidata et al...

It is all well and good that there is this grant that comes to a conclusion but the fact of the matter is that people aim to start removing items because they deem the quality of the scholarly works these items represent too poor. This is possible because they see no application for that data, 

I am adding papers and books all the time, the books I link to Open Library because our objective should be for people to read. I add citations to scholarly works when I consider those works important, I disambiguate author strings to give more weight to both the author and the works involved. It has value because PigsOnTheWing and others are linking references in Wikipedia to Wikidata items. It has value because Scholia templates refers to these authors, scholarly works and, Scholia knows in what Wikipedia articles a paper is used. The tools that were available to do all this have largely gone away so effectively we are in a poor state to improve on what there is.

But in the final analysis WikiCite is dead when once poked it does not react. Its contributions have no value when that value is not understood and removed.  A Dutch poet, Lucebert, famously said "alles van waarde is weerloos". Liam I see the work that I put in come to naught and you accuse me of bringing a message that "is not appropriate". What I see is that the message of WikiCite is being lost and I now understand that WikiCite is coming to an end as well. What we need is to expand on what is achieved, not as a stamp collection but as a data collection that with associated tools invites people to do further reading. A collection that makes plain that the science on a subject has aged. Our work needs an easy application. To do that we have to collaborate with partners like Open Library, Internet Archive, ORCiD, Crossref and to be honest I think we suck at it.

Thank you,

         GerardM

[Mark Graham]

For me a significant part of the future of “Wikicite” is this project: https://meta.wikimedia.org/wiki/WikiCite/Shared_Citations

And, if all that those who developed projects, and services, under the umbrella of “Wikicite” for the past few years, have helped contribute to a process that got us to a space where the value of, and a plan for, a Shared Citations project is clear, than I will call it all a success.

For me Wikicite has always been about expressions of shared understandings of the importance and usefulness of citations, to Wikipedia.  Including the ability to access the cited material with a click and to study and otherwise analyze those citations.  It has been a support structure for experiments and collaborations.  And that has been of value in and of itself.

In the meantime, the Internet Archive’s Turn All References Blue project will continue to add links to Wikipedia articles, connecting to archived Web pages and Papers and digitized books… efforts that I have always thought were inspired by and part of other “Wikicite” efforts, even it they were not focused on Wikidata.  I guess, in that way, Wikicite is what any of us want to make it.  And, I fully expect the outcome of those efforts 5 years from now will surprise most of us.  I think the best days of Wikicite are ahead of us.

I will also share that, from the Internet Archive’s perspective (including Open Library) collaborations with other Wikipedians and the Wikimedia Foundation(s) has never been better.

Now… as for the use of Telegram… well… that is another story altogether ::-)

Long live Wikicite!

- Mark Graham
Director, the Wayback Machine @ the Internet Archive

---
To unsubscribe from this group and stop receiving emails from it, send an email to wikicite-discu...@wikimedia.org.

[Andra Waagmeester]

 The problem with both Wikidata and Wikicite hiding on Telegram is that for reasons they deem to be significant we have lost the conversation and lost their conversation with others. There is also a Facebook environment where people talk about Wikidata et al...

 Don't forget, IRC, Reddit, Twitter, stackoverflow, etc. The problem with those (great) platforms is that the messages expressed there are not documented nor do they persist, which makes it impossible to cite (pun intended). Messages expressed in the socials should be taken for what they are: Gossip. There are so many platforms that any message is easily missed. 

 

But in the final analysis WikiCite is dead when once poked it does not react. Its contributions have no value when that value is not understood and removed. 

If people are removing properly referenced statements because they don't like it, they are vandals. 

 

A Dutch poet, Lucebert, famously said "alles van waarde is weerloos". Liam I see the work that I put in come to naught and you accuse me of bringing a message that "is not appropriate".

To stick with the Dutch, there is also Frits Bolkestein, who once said that a country whose history is boring, is a lucky country. The fact that it is not being discussed or actively marketed, does not imply that it is not being used. You could easily argue that it is relatively quiet on the PR side because people are busy working on adding, reusing, and building on the Wikicite corpus. 

 

What I see is that the message of WikiCite is being lost and I now understand that WikiCite is coming to an end as well. What we need is to expand on what is achieved, not as a stamp collection but as a data collection that with associated tools invites people to do further reading. A collection that makes plain that the science on a subject has aged. Our work needs an easy application. To do that we have to collaborate with partners like Open Library, Internet Archive, ORCiD, Crossref and to be honest I think we suck at it.

I honestly don't think "we" suck at it. On the contrary, another "we",  just published a paper [1]  where we stipulate the steps involved in getting knowledge into Wikidata, where  Wikicite, in all its facets as nicely described by Liam, provide the backbone of the provenance of that work. To give you an idea of the timeframe. It took us almost a year to not only do the work but also get this paper out. It was hard work and when doing so, we easily miss fluid discussions on telegram and as such don't respond to those pokes your mention.  Wording it like "we suck" is just an insult. The fact that some don't see that in the chat sections of the socials, is on them and not on us who are doing a lot of this in the margins of our spare time.  

 

[1] https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-020-00940-y

[Voß, Jakob]

Hi Gerard,

Sorry for the disappointment and waiting. As far as I understand your concern by large parts about the goal of combining all publications in a public database. From my point of view this is only one aspect of WikiCite and I doubt that it's fully feasible. Anyway, another thing: you wrote:

> We do not know all the national ministers for the twenty first century and obviously not for the twentieth century of the African countries.

By accident I'm about to import information about African ministers to Wikidata. The project slept for some month but I just started to get back to it this weekend. The data contains 16.000 positions between 1941 and 2005:

https://www.wikidata.org/wiki/Wikidata:WikiProject_Africa/Cabinets​. 

Cheers,
Jakob

[Luca Martinelli]

Il giorno lun 15 feb 2021 alle ore 13:46 Voß, Jakob
<Jakob...@gbv.de> ha scritto:

> By accident I'm about to import information about African ministers to Wikidata. The project slept for some month but I just started to get back to it this weekend. The data contains 16.000 positions between 1941 and 2005:
> https://www.wikidata.org/wiki/Wikidata:WikiProject_Africa/Cabinets.

Super-wow! I've done my bit for the Italian cabinets, maybe we want to
coordinate better at Project:Politics!

(Whoops, sorry for the OT)

--
Luca "Sannita" Martinelli
http://it.wikipedia.org/wiki/Utente:Sannita

[James Hare]

On Sun, Feb 14, 2021 at 6:44 AM Gerard Meijssen <gerard....@gmail.com> wrote:

• Magnus rewrote many of his tools in RUST, including the SourceMD tooks, it made no difference for the community

I don't want this to go unappreciated. A few months ago I used QuickStatements for the first time in years and I was very impressed with how fast it was. Considering how essential QuickStatements is, I am greatly appreciative of Magnus' work making QuickStatements more robust and scalable.