The Wikimedia Foundation Executive Director (ED) and Chief Executive Officer Katherine Maher announced February 4 that she would be leaving the WMF on April 15. She has been with the WMF seven years, first as chief communications officer, then as interim ED in March 2016, and as ED in June 2016.

She became ED at a very difficult time for the movement, reported on in The Signpost as The WMF's age of discontent. Many in the community and WMF staff were outraged by a search engine project known as the knowledge engine. Doc James was voted off the Board of Trustees by the Board itself. Maher's predecessor, Lila Tretikov, resigned under pressure. A new trustee resigned within weeks of his appointment. Following Maher's appointment the community, staff, and Board experienced a period of peace and cooperation for about three years.

Maher's announcement was surprising since just the day before she announced the approval of the Universal Code of Conduct by the Board of Trustees. Together with the 2030 strategy recommendations, the UCC may be the chief achievement of her term as ED. Both still require implementation or enforcement, which may be just as difficult to achieve as the agreement on the policy and goals.

The WMF described the major improvements reached during Maher's term as:

• Increasing global reach and readership
• Reversing editor decline, and generating editor growth
• Increasing editor diversity
• Growing a global movement

Maher was an unapologetic advocate for a diverse community. She told The Signpost in a 2019 interview:

Diversity is baked into our vision statement: the sum of all knowledge, every single human being. And feminism is a foundational part of diversity: if we’re talking about every single human being, we need to be talking about every single human being, including women and non-binary people. So, not only is this part of my values, it’s absolutely part of the Foundation’s mission.

Relations among the community, WMF staff, and the Board of Trustees became contentious at times. A consultant's report published in 2020, meant to better organize WMF governance, recommended that communications between the Board and the ED should be strengthened and that the ED's office should be better staffed to handle the many challenges.

In a note to Wikipedians on diff, Maher said "I’m going to take a break, and a research fellowship, as a place to think about what’s next." Axios stated that she'll be moving to the United States east coast.

General Counsel Amanda Keton, Chief of Talent and Culture Robyn Arville, and Chief Financial Officer Jaime Villagomez will act as the executive transition team. The Transition Committee leading the search for Maher's replacement includes four trustees: Dariusz Jemielniak, Tanya Capuano, Raju Narisetti, and Board Chair María Sefidari. Viewcrest Advisors will assist the committee's global search.

See this formal announcement.

The Signpost wishes Katherine all the best. – S

TonyBallioni opened a discussion at Wikipedia:Requests for comment/Desysop Policy (2021) on February 20 to streamline a process to remove administrators, driven more directly by the community. It has 126-80 support as of February 27.

Current procedures to remove an administrator on the English Wikipedia (desysop) are outlined at WP:Administrators § Review and removal of adminship and do not include a direct path for the editors in the community to remove an administrator. Not counting procedural and reversible desysop due to inactivity, or desysops initiated by WMF which are counted in single digits, desysop virtually always happens through the admin's own request or by Arbcom action. Currently 146 of 512 currently-active admins, listed at Category:Wikipedia administrators open to recall, have voluntarily specified their own ad-hoc criteria and procedures for desysop that they have pledged to follow. Thus over 70% of active admins are only removable by themselves or by Arbcom.

A new process has been a recurring topic on the English Wikipedia, with discussions going back to 2004 listed at Wikipedia:Requests for de-adminship. A 2009 discussion to formalize the process, Wikipedia:WikiProject Administrator/Admin Recall, resulted in a 2010 RfC which failed: 167 for, 190 against; and a 2019 RfC failed to result in a procedure [that] gathered enough support to be favored by a clear consensus.

Conditions outlined by the current RfC in order to initiate a desysop request include (condensed for The Signpost):

• A recent AN or ANI discussion where the closing statement indicates that there was consensus that the administrator behaved inappropriately
• Endorsement by a qualifying number of editors, including three admins

The desysop will be performed by a bureaucrat if there is 60% support for removal at the end of a 7-day voting period. The word "vote" is not used in the initial RfC statement and it is being debated whether the proposed process will use a straight-up vote or a discussion to determine the outcome. – B

• New administrators: The Signpost welcomes the English Wikipedia's newest administrator, TJMSmith.

Several billionaires – or more likely their paid representatives – appear to have edited Wikipedia according to an article in The Wall Street Journal^[1] and in an investigation published in The Signpost. Two alleged sex offenders, Jeffrey Epstein and Ghislaine Maxwell, also appear to have edited Wikipedia.

Do these types of editors interact on Wikipedia? Do they have edit disputes with one another? The Signpost investigates these questions in the case of a "billionaire battle"^[2] between hedge fund manager, Louis Bacon, whose net worth was $1.5 billion in 2020 according to Forbes, and Peter Nygard, a fashion executive who has been indicted on nine charges in New York, which include sex trafficking, money laundering, and racketeering.

Nygard never quite made it into the list of billionaires on Forbes, but his net worth was estimated at $750 million in 2014.^[3][4] Today he is being held in Winnipeg, Manitoba, Canada, for extradition to the United States. He has been unable to raise bail and remains in jail.^[5]

Nygard's lawyer has denied all the charges against him. The Signpost reminds our readers that he should be considered innocent until any charges are proven in a court of law. We also remind you that the identities of Wikipedia editors can never be completely proven – even if they seem to have identified themselves. They may be spoofing or "Joe jobbing" in order to embarrass other people.

Nygard emigrated with his parents from Finland to Canada when he was ten years old and entered the clothing business after graduating from university. He bought into a small firm in Winnipeg and then bought out his partners. His group of companies, headed by Nygard International, grew to 165 retail outlets in the US and Canada^[6] and also sold through Dillard's and Walmart in the US.

He owned a home in California and in 1987 bought property in an exclusive gated community, Lyford Cay in the Bahamas. He built an immense Mayan-themed complex on the six acres of prime beach-front property which was later featured on the television show Lifestyles of the Rich and Famous.^[2]

Just over a year ago he might have still thought that he was living his dream life, as he flew to his various businesses and homes in a private Boeing 727 jet with an entourage of young women and under-age girls. The only fly in the ointment seemed to be a feud^[2] with his neighbor in the Bahamas, billionaire Louis Bacon.

An argument over a shared driveway is reported to have started the feud, which grew into a legal dispute about Nygard's use of dredging to expand his land along the coast. Other accusations followed. Nygard accused Bacon of blasting out Nygard's parties with industrial-grade loudspeakers. Nygard accused Bacon of insider trading, arson, and even murder. He organized a demonstration against Bacon’s supposed membership in the Ku Klux Klan. Bacon, in turn, accused Nygard of planning his murder.^[2][4]

How do the very rich deal with disinformation being spread about them? One method is to file a lawsuit. According to The New York Times, the two filed 25 lawsuits against each other in five jurisdictions.^[4] Bacon also took legal action in the UK against the Wikimedia Foundation to force the WMF to help identify Wikipedia editors who he believed were defaming him. He won a UK court order, but could not get it enforced in the US. The WMF likely does not have the identifying information in any case.^[7]

On February 13, 2020, Nygard was sued by 10 "Jane Does" in civil court in New York, alleging rape. The lawsuit was widely viewed as the work of Bacon.^[4] Several dozen more accusers have since claimed that Nygard raped them. The rapes allegedly occurred from 1979 through 2020. Another alleged rape, reported in the Winnipeg Free Press, was said to have taken place around 1960 when Nygard and the unnamed woman were in high school.^[8] Several plaintiffs say they were minors when the alleged rapes occurred.

On February 25, 2020, the FBI raided Nygard's Times Square offices in Manhattan. Nygard resigned the same day from his own companies,^[8] most of which soon filed for bankruptcy. By May they were being sold off in pieces to satisfy a mere $25-million debt. US federal prosecutors, on December 15 unsealed a nine count indictment against Nygard in the Southern District of New York,^[12][13] including allegations of sex trafficking, multiple cases of coerced sex with underage girls, coerced participation in orgies, racketeering, and money laundering. He was arrested in Winnipeg and held for extradition to the US. He has been unable to get bail and remains in jail. At age 79, his lawyers claim that he is broke, sick and dying.^[14]

Four accusers have been interviewed and shown in documentaries on the Canadian Broadcasting Company. They describe forcible rape that appears to have been planned beforehand. They speak at length and their identities are not hidden. The CBC also shows video from Nygard's personal videographer, who says that he has about 1,000 hours of video of Nygard.

One pro-Nygard editor, User:NYGARD International, only edited the Peter Nygard article and only on October 30, 2007, making eights edits before they were quickly blocked for promotional editing. One of their edits completely rewrote the article with 3,116 words of promotional material starting

Most people talk about Peter J. Nygård, Chairman of NYGÅRD(International), in terms of his classic rags-to-riches story – the Finnish immigrants' son, who stitched up an empire out of women's clothing and is now the quintessential self-made man. This story overlooks another side of Peter Nygård – a hard driven, demanding and subjective man who has created a standard of excellence for the Canadian Women's Fashion Industry, whose label is the #1 recognized label in the Canadian marketplace, and whose signature is recognized in fashion centres across the globe.`

While promotional edits are common on Wikipedia, few, if any, are so open, so non-encyclopedic, and so poorly written. The real disinformation about the Nygard/Bacon feud though began three years later after the CBC broadcast a documentary about Nygard's alleged harassment of his employees. Nygard responded in the courts by filing a criminal libel suit. He apparently responded on Wikipedia by attacking Bacon through several sockpuppet accounts and trying to delete every mention of the CBC documentary.

Bacon indicated that he wanted to start defamation proceedings against some Wikipedia editors by obtaining court orders from the high court in London to force three website owners – the WMF, WordPress, and The Denver Post – to disclose the identities of contributors who Bacon believed defamed him, according to The Guardian.^[7] The WMF could not be forced to provide the information without the intervention of a US court, and likely did not ever possess the required information. The Guardian reported the online names of the accused defamers as "gotbacon" and "TCasey82".

"Gotbacon" is not a registered Wikipedia username, but likely refers to the WordPress blog gotbacon which published between December 2010 and May 2011. It republished anti-Bacon material from little known websites, commented on Bacon and complained that the Wikipedia article on Louis Bacon was whitewashed. It was cited as a reference in 2011 by User:Crinock in the Louis Bacon article. Crinock was indefinitely blocked as a sockpuppet of RK Drollinger in the same year. Another blocked sock of RK Drollinger, User:Rosi.anastasova, made five of their nine edits to the Louis Bacon article, deleting material reported in the 2010 CBC documentary.

User:TCasev82 was a single-purpose account (SPA) who made all 33 of their edits between May 2010 to February 2011 to the Louis Bacon article. They included anti-Bacon material and inserted an external link to the gotbacon blog.

User:Lbninternational made all 17 of their edits on November 15, 2011 to the Louis Bacon article before they were warned for a biography of living persons policy violation: "Please stop trying to blackball Louis Bacon". They were then reported to WP:ANI where Louis Bacon's legal action against the WMF was discussed. Lbninternational has not edited since that time. One of their edits read "Although Lpuis Bacon [sic] is reportedly the 655th richest person in the world, there have been some bizarre stories linked to him such as this one in Business Insider" while linking to a story on an alleged murder.

Did Louis Bacon or his representatives respond to these edits on Wikipedia? Complaints sent to Wikipedia via the Volunteer Response Team ticket request system (better known as OTRS) are essentially private and The Signpost does not have access to them, but it appears that Bacon complained about these and other edits through OTRS in 2011 and again in 2019. These requests appear to have been handled by administrators and experienced volunteer editors according to policy, mostly in favor of Bacon's position. In 2011 there were fewer reliable sources reporting on Nygard and regular Wikipedia editors took longer to decide matters, but most results favored Bacon.

In 2018 and 2019 User:Candor777 made a total of nine edits to the Nygard and Bacon articles, all of which favored Nygard. They edit warred to remove documented accusations against Nygard. They were warned about our rules on conflict of interest and paid editing. Their edits were reverted and they stopped editing.

Two very rich men, one a billionaire, the other accused of multiple forcible rapes and multiple rapes of minors, may have battled on the pages of Wikipedia leading to a court ordering the WMF to identify our editors for inclusion in a defamation suit. This possibility should appall all Wikipedia readers and all Wikipedia editors. But did it happen that way?

Peter Nygard and Louis Bacon did conduct an off-wiki "billionaire battle" campaign of accusations and lawsuits against each other for a decade. Did the battle spill over onto Wikipedia? Bacon seemed to think so, having sought the court order to identify Wikipedia editors, and apparently complaining twice through OTRS about false information in Wikipedia articles. The Signpost, however, was unable to identify any article edits made by Bacon or his representatives.

Since the identities of Wikipedia editors can not be completely proven by their editing histories, The Signpost can only say that there is a strong likelihood that Nygard or his representatives edited Wikipedia. User:NYGARD International was quickly blocked for their promotional editing to the Peter Nygard article. Gotbacon, a WordPress blog that attacked Bacon and complained that the Bacon article was whitewashed, connected several apparent Nygard editors, including User:TCasey82 and User:Crinock, both single-purpose accounts linked to the blog. Crinock was blocked for sockpuppeting connecting them to another sockpuppet account User:Rosi.anastasova, who also edited the article. Just before he stopped editing in 2011, SPA User:Lbninternational was warned by an administrator to "stop trying to blackball Louis Bacon". Eight years later, SPA User:Candor777 quit editing after they were warned about their editing.

Yes, it appears that Nygard or his representatives were continuing his off-wiki feud on Wikipedia.

The Wikimedia Foundation is taking away your freedom to elect community representatives to their board.

In 2017 some 5,500 contributors to Wikimedia projects elected three members to the Board of Trustees of the Wikimedia Foundation for a three year term. Their mandate ended nearly a year ago. The Board decided to extend their terms for a year. The ongoing pandemic was an excuse to postpone an election, yet the Board of Trustees launched a very intensive consultation with the community. This consultation is probably more time- and energy-consuming than an actual election.

They want to discuss changes to the way Trustees are selected. They aim at greater diversity. They have concerns about skills and experiences of community elected board members.

This consultation isn't the only ongoing consultation with the communities. There are a whole range of Global Conversations about implementation of the recommendations of the Wikimedia 2030 Movement Strategy process. And some other consultations are also running. One of the strategy recommendations – which is priority 1 – is to ensure equity in decision making. To include underrepresented voices in governance and to assure diversity, it was recommended to have a Global Council in the future. The Global Council would be a large body, representing the diversity of the movement.

There will be a Global Council at the earliest at the end of Wikimania 2022. Depending on your point of view that is either very far away, or pretty soon. In the interim, there will be an Interim Global Council to draft a Movement Charter and oversee the implementation of the strategy recommendations. The current timeline projects having an Interim Global Council seated in April or May 2021. That is pretty soon, and probably at the same time or even before new community representatives will be elected to the Board of Trustees.

The movement has recognized and acknowledged a lack of diversity of our representation in governing bodies over the past years, and has resolved this issue by proposing an (Interim) Global Council (IGC). The current call for feedback about community seats is not only badly timed – in the midst of a pandemic, on top of multiple other consultations, but also tries to seek to solve a problem – diversity – which has already been solved by having an IGC soon. And the board failed to integrate this consultation within the overall framework of conversations about implementation of strategy recommendations. Just like they failed at integrating the branding discussion within the strategy process.

The other part of the 'problem' the Board seeks to solve is with respect to skills and experiences. Five trustees on the board are selected by communities and affiliates from volunteers who contribute to Wikimedia projects. Those volunteers have multiple years of experience within the Wikimedia Movement, are skilled in editing Wikimedia projects, and skilled in contributing in other ways to the Wikimedia Movement. Most notably, those volunteers have taken initiatives both online and offline to organize projects and programs. It looks like the current board doesn't value these skills and experience as relevant to governing the Wikimedia Foundation.

Half the Board of Trustees of the Wikimedia Foundation consist of appointed trustees chosen for their specific skills and experiences. Most of them aren't contributors to the Wikimedia projects. Most of them had no idea how Wikimedia projects function and operate, and are unaware how Wikimedia volunteers have organized themselves online and offline over the past.

Most notably there is a total lack of diversity among appointed trustees, most of them being US residents. The five volunteers on the board all speak different native tongues, and are all born in different countries, originating from four different parts of the world. The current diversity among community and affiliate seats is at a maximum given the number of available seats. If there is a diversity problem on the board, then it is within the appointed seats, not within the community seats.

In 2019 a woman from Ukraine (Eastern Europe) and a woman from Israel (Middle-East) were selected by affiliates to become board members. The call for feedback states as problem the election process favors men from the US and Western Europe. The latest election outcome fully contradicts this part of the problem statement.

The Board amended the bylaws in 2020. They proposed expanding the board from 10 to 16 seats. There will be even more appointed seats, not selected by contributors to Wikimedia projects, in the future. The call for feedback is restricted to community seats. The Board didn't ask the community for advice how to fill in the extra appointed seats they created.

What the Board of the Wikimedia Foundation actually does is mostly boring stuff. They have to read piles of paper, approve annual plans and annual budgets, review financial reports and go through tons of legal stuff. The board also does hire and fire the CEO or Executive Director of the Wikimedia Foundation, the person who actually leads the 450+ staff of the Wikimedia Foundation on a daily basis.

Apparently the Board sees an urgent need to hear from a voice from Africa. They created three extra appointed seats, and no one is blocking the Board to handpick someone from Africa to become one of the Trustees. And maybe another one from South-East Asia. Apparently they want someone on their board who has previous experience as CEO or Executive Director of a non-profit of comparable size working in the same field as the Wikimedia Foundation. No one is blocking the Board appointing Katherine Maher to the Board, as she has announced her departure as CEO/ED of the Wikimedia Foundation by April 15. All in all, this leaves ample room for the community and affiliates to select their representatives through their preferred method, that is by having a free and open election.

Apparently the Board has a dislike for elections, and wants to get rid of the system of election by the community and the affiliates of Trustees. In preparing this article I have reached out to all board members, asking them whether they would prefer appointment of parliamentarians rather than electing them by the people in their country. Some of the board members engaged in email conversation, and I raised more questions. The ensuing conversation stopped soon after the moment someone realized they were stepping out of line, and ranks were closed. I could get a quote from an official voice from the board. That quote ended up being 462 words long, not answering any of my questions directly.

The call for feedback about community seats proves the Board of Trustees is not capable of governing the Wikimedia movement, and the call for feedback on community seats is a distraction from the conversation on implementing the Wikimedia movement strategy recommendations. The Wikimedia movement has an urgent need of a Global Council inclusive of all Wikimedia project wiki communities.

Whether or not Wikimedia project wiki communities actually do support the creation of an (Interim) Global Council and drafting a Movement Charter, which are highly preferred by some people from Wikimedia affiliates, is a topic for a Signpost article in the future.

Inside the 'Wikipedia of Maps', Tensions Grow Over Corporate Influence (Bloomberg) examines "digital gentrification" of OpenStreetMap which between 2015 and 2018 saw a sixfold increase in features edited by corporations, led by Apple. Lyft, Facebook, the International Red Cross, the U.N., the government of Nepal and Pokémon Go all depend on its data. In turn, "hundreds of millions of monthly users" depend on the organizations' use of the data. Commercial firms are protecting their investment in the data by editing and contributing more data. One startup, Mapbox, has raised $200 million to develop ways to format and transfer OSM data to its customers. Is this just another example of the benefits of crowdsourcing, or just another corporate takeover of the commons? "Hobby mappers" are worried that corporate representatives will be elected to the site's governance positions. Frederik Ramm, an OSM volunteer and consultant states "These companies don’t map for the same reasons we do, and because of that, I question deeply if our goals can align."

Wikipedia's Sprawling, Awe-Inspiring Coverage of the Pandemic in The New Republic joins dozens, likely hundreds, of articles on Wikipedia's coverage of the COVID pandemic, the first of which was published 384 days ago in Wired by Omer Benjakob.

Though this ground has been thoroughly plowed, Shaan Sachdev in TNR reports with a different angle. He states that with 86 million pageviews "the Covid-19 pandemic (article) is in a two-way tie for the thirty-fourth most viewed Wikipedia article, ranking alongside Miley Cyrus. It isn’t far behind thirty-first place, which is currently a three-way tie between Taylor Swift, Star Wars, and China."

Some of the other articles that link to COVID-19 are listed (in no special order) as: Mink, Racism in China, Royal Australian Navy, Graffiti, Cockfight, and Ricky Martin. Ultimately almost all articles of this genre focus on Wikipedia's volunteer editors, as they should. Editors included in this article include Andrew Lih, Netha Hussain and Liam Wyatt.

Fox News interviewed Larry Sanger in Inside Wikipedia's leftist bias: socialism pages whitewashed, communist atrocities buried. Sanger's views on Wikipedia's "leftist bias" and chaotic governance have been well-known since 2007. Over the years he's made a few good points, for example his claims that Wikimedia Commons contained some child pornography, but Fox's claim that he told them that "many Wikipedia pages have become merely left-wing advocacy essays" is exaggerated. Bias is difficult to define without having a grasp of the range of views typically accepted within a population. Fox's hard right views do not define bias within the US. Wikipedia's editors come from many countries beyond the US that generally have views to the left of the US. Fox does score points when looking at Wikipedia's coverage of Communism. Coverage of genocide by Communist governments is all but missing on Wikipedia – even when we call it "mass killings".

• Fox gives a specific example of possible bias in another article Wikipedia restricts edits to Lincoln Project page amid sex scandal, claiming 'vandalism, as if there was no vandalism to the article.
• The Daily Mail repackages Fox's reporting on Sanger in Wikipedia co-founder Larry Sanger slams the site's left wing 'woke' bias and claims its days of 'neutrality are long gone'. Needless to say, Fox's original is better.
• Co-founder says Wikipedia's neutrality 'long gone,' cites leftist bias in The Washington Times says "Sanger tweeted that leftist activists gradually move in 'to take control of any influential institution not explicitly conservative … and they just work harder, and in more subtle ways, on the ones that are explicitly conservative.' " The attribution to Sanger is questionable and makes him sound like a conspiracy theorist. Folks who question why these sources aren't considered reliable by Wikipedians need only read these articles.

Wikipedia’s political science coverage is biased. I tried to fix it. in The Washington Post. Samuel Baltz, a Ph.D. candidate at the U. of Michigan in political science and computing, spent a year trying to correct the biases he sees in Wikipedia articles on political science.

• Why Is This Flower on Wikipedia Suddenly Getting 90 Million Hits Per Day? published on Vice.

• Wikidata for journalism and beyond: DataJournalism.com from publisher European Journalism Centre and author Monika Sengul-Jones report Wikimedia editors' views on how their Wikidata editing improves the media ecosystem for good journalism.
• Science fiction in Wikipedia's style: The science fiction community of the SCP Foundation is a non-Wikimedia wiki network that's growing in popularity. Wikipedia science fiction fans should like the recently published "Lena", in which the author qntm presents a story in the format of a Wikipedia article written some time after 2073. The article is roughly about a brain scan and the "first immortal". It is copyrighted so cannot be used to update Wikipedia in real time.
• Wikimedia CEO: Technology could be used for great harm: Katherine Maher has a 25 minute interview with Al Jazeera.
• Happy hump day: A Wikipedia page appears as an easter egg ... or trolling? ... from Henry Cavill. (Yahoo! News)
• Turkey's pro-gov't media still trying to get Wikipedia banned by Wikipedian John Lubbock. "Turkey's pro-government media have not given up accusing Wikipedia of various crimes."
• Click! on the BBC includes an interview with Jimmy Wales.
• Editathon on EDSA Revolution-related topics hosted by Wiki Society of the Philippines (Philippine Daily Inquirer)
• IDF Asks Wikipedia to Edit 'Hezbollah' Entry to Reflect Terror Designation (The Jerusalem Post)
• Myanmar blocks Wikipedia in all languages (India TV)
• Someone at Apple is editing ZZ Top's Wikipedia articles (iMore)

• Digital Unionism is Generally is Quite New interview on WikiProject Labor's edit-a-thon. Read their report back. (netzpolitik.org)

• World in Progress: Wikipedia's past, present and future - extended interview: (audio) Deutsche Welle interviews Dr. Bernie Hogan from the Oxford Internet Institute.
• An Oral History of Wikipedia, the Web’s Encyclopedia: Tom Roston in OneZero conducts extensive interviews including Jimmy Wales, Larry Sanger, Katherine Maher, Andrew Lih, Britta Gustafson, Magnus Manske and Jeremy Rosenfeld.
• The Economist: Readers responding to The Economist's extensive coverage of Wikipedia's 20th birthday had very interesting comments, including these two:

Anyone who has ever tried to correct errors in a Wikipedia entry, only to find them repeatedly reinserted by other contributors with a competing agenda, will attest to the site's unreliability. Even counting on Wikipedia as a repository of basic information, such as names, dates and places, is a crap shoot. Perhaps the vast majority of its articles are indeed accurate, but which ones constitute that majority, and at what point in time? Literally no one knows; it has become so vast that moderating its millions of entries in any comprehensive way would be impossible. This is how the site is designed to work.

— b.e.

Wikipedia may be unique online because it "sells no advertising". However, it does provide a free platform for companies to display their corporate messages, written by their marketing departments. Are these true statements vetted by Wikipedia? No. Wikipedia also has a devil’s bargain with Google. No matter what you search for on Google, from "cats" to "Catullus", Wikipedia is positioned first. If you do enter "cats", to learn about the animal, you get what reads like a Wikipedia advertorial for the movie "Cats".

— p.t.

A monthly overview of recent academic research about Wikipedia and other Wikimedia projects, also published as the Wikimedia Research Newsletter.

Reviewed by Tilman Bayer

Flashcards are a popular method for memorizing information. A paper^[1] by six Zurich-based researchers, presented earlier this month at the annual AAAI conference, describes a tool to automatically extract flashcards from Wikipedia articles, aiming "to make independent education more attractive to a broader audience."

A proof-of-concept version is available online, with results available for export in a format that can be used with the popular flashcard software Anki. User can choose from four different variants based on either the entire Wikipedia article or just its introductory section.

The researchers emphasize that "generating meaningful flashcards from an arbitrary piece of text is not a trivial problem" (also concerning the computational effort), and that there is currently no single model that can do this. They separate the task into four stages, each making use of existing NLP techniques:

• summarization, to first extract the most relevant information from Wikipedia (the user can also choose to have this step skipped and instead generate flashcards based on the full text)
• answer identification, where a model extracts answer statements from a given sentence based on context information from the surrounding paragraph
• question generation, where a model constructs a question from the statement generated in the previous step, again taking context information from the surrounding paragraph into account
• To improve quality, these are followed by a final filtering step, where a question-answering model tries to reconstruct the answer based on the paragraph from which the question was extracted, and the generated flashcard is discarded if the reconstructed answer does not overlap enough with the pre-generated answer.

Apart from evaluating the results using quantitative text measures, the researchers also conducted a user study to compare the output of their tool to human-generated flashcards from two topic areas, geography and history, rated by helpfulness, comprehensibility and perceived correctness. The "results show that in the case of geography there is no statistically meaningful difference between human-created and our cards for either of the three aspects. For history, the difference for helpfulness and comprehensibility is statistically significant (p < 0.01), with human cards being marginally better than our cards. Neither category revealed a statistically significant difference in perceived correctness." (However, the sample was rather small, with 50 Mechanical Turk users split into two groups for geography and history.)

A quick test of the tool with the article Wikipedia (introduction only) yielded the following result (text reproduced without changes):

Question: What does Wikipedia use to maintain it's [sic] content?

Question: In 2021, where was Wikipedia ranked?

Question: What language was Wikipedia initially available in?

Question: How many articles are in English version of Wikipedia [sic] as of February 2021?

Question: Who hosts Wikipedia?

Question: Whose vision did Time magazine believe made Wikipedia the best encyclopedia in the world?

Question: What is a systemic bias on Wikipedia?

Question: What did Wikipedia receive praise for in the 2010s?

Question: What two social media sites announced in 2018 that they would help users detect fake news by suggesting links to related Wikipedia articles?

• See the page of the monthly Wikimedia Research Showcase for videos and slides of past presentations.
• @WikiResearch, the Twitter feed associated with this monthly research update, celebrated its ninth anniversary today. Over the past 9 years, we have shared on average 1.9 tweets per day about Wikimedia-related research. The feed is also available in syndicated form on Facebook and Mastodon.

Other recent publications that could not be covered in time for this issue include the items listed below. Contributions, whether reviewing or summarizing newly published research, are always welcome.

Compiled by Tilman Bayer and Miriam Redi

A paper in New Media & Society^[2] argues that

"[...] an 'implicit feudalism' informs the available options for community management on the dominant platforms for online communities. It is a pattern that grants user-administrators absolutist reign over their fiefdoms, with competition among them as the primary mechanism for quality control, typically under rules set by platform companies.

[...] the online encyclopedia Wikipedia operates through a sophisticated democracy among active volunteers. Wikipedia also possesses a widely acknowledged benevolent dictator in the person of founder Jimmy Wales [...] Implicit feudalism has reigned over the dominant platforms for online communities so far, from the early BBSes to AI-enabled Facebook Groups. Peer-production practices surrounding free/open-source software and Wikipedia also exhibit it.

[....] The feudal pattern has by and large been written into the default behaviors of online-community platforms. Exceptions like Wikipedia and Debian have required considerable, intentional effort to counteract the implicit feudalism of their tools’ defaults."

From the abstract:^[3]

"Using a novel technique, a massive database of qualitatively described citations, and machine learning algorithms, we analyzed 1 923 575 Wikipedia articles which cited a total of 824 298 scientific articles in our database and found that most scientific articles cited by Wikipedia articles are uncited or untested by subsequent studies, and the remainder show a wide variability in contradicting or supporting evidence. Additionally, we analyzed 51 804 643 scientific articles from journals indexed in the Web of Science and found that similarly most were uncited or untested by subsequent studies, while the remainder show a wide variability in contradicting or supporting evidence."

From the abstract:^[4]

"Collecting supporting evidence from large corpora of text (e.g., Wikipedia) is of great challenge for open-domain Question Answering (QA). Especially, for multi-hop open-domain QA, scattered evidence pieces are required to be gathered together to support the answer extraction. In this paper, we propose a new retrieval target, hop, to collect the hidden reasoning evidence from Wikipedia for complex question answering. Specifically, the hop in this paper is defined as the combination of a hyperlink and the corresponding outbound link document."

(See also the above review of the "WikiFlash" paper presented at the same conference)

From the abstract:^[5]

"... we employ question answering and entity summarization as extrinsic use cases for a longitudinal study of the progress of KB coverage. Our analysis shows a near-continuous improvement of two popular KBs, DBpedia and Wikidata, over the last 19 years, with little signs of flattening out or leveling off."

See also the video recording of a talk by the authors at Wikidata Workshop 2020.

Presented at the ACM Special Interest Group on Information Retrieval (SIGIR) forum last December, this paper^[6] found that the majority of Question Answering (QA) datasets are based on Wikipedia data.

From the "Evaluation" section of an AAAI'21 paper titled "Identifying Used Methods and Datasets in Scientific Publications":^[7]

"Figure 4c shows the absolute amount of publications for the top four extracted datasets. [...] Another trend is visible for Wikipedia, which has become popular in research on knowledge representation and natural language processing."

The contributions of this paper^[8] include

"a hub of pre-indexed Wikipedia [dumps, of the English and Chinese language versions] at different years with different ranking algorithms as public APIs or cached results". The authors note that "Opendomain QA datasets are collected at different time, making [them depend] on different versions of Wikipedia as the correct knowledge source. [...] Our experiments found that a system’s performance can vary greatly when using the wrong version of Wikipedia. Moreover, indexing the entire Wikipedia with neural methods is expensive, so it is hard for researchers to utilize others’ new rankers in their future research."

This preprint^[9] includes a dataset consisting of 17 conspiracy theory topics from Wikipedia (including e.g. the articles Death of Marilyn Monroe, Men in black, Sandy Hook school shooting) and comes with a content warning ("Note: This paper contains examples of potentially offensive conspiracy theory text").

From the abstract:^[10]

"[We analyze] the Wikipedia edit history to see how spontaneous individual editors are in initiating bursty periods of editing, i.e., spontaneous burstiness, and to what extent individual behaviors are driven by interaction with other editors in those periods, i.e., interaction-driven burstiness. We quantify the degree of initiative (DOI) of an editor of interest in each Wikipedia article by using the statistics of bursty periods containing the editor's edits. The integrated value of the DOI over all relevant timescales reveals which is dominant between spontaneous and interaction-driven burstiness. We empirically find that this value tends to be larger for weaker temporal correlations in the editor's editing behavior and/or stronger editorial correlations. These empirical findings are successfully confirmed by deriving an analytic form of the DOI from a model capturing the essential features of the edit sequence."

(See also our earlier coverage of research on editors' burstiness)

I do feel my attempts to have relevant titles related to monthly holidays is getting increasingly desperate, but, well, here we are. It's either that or we go with the list off a few random things featured in the list, which seems fine on occasion, but does get a bit repetitive if done every month. And I struggle enough to avoid getting my biases into this when it's just down to which articles get images - for the record, the rule is: if it has a good-quality, freely-licensed image, directly relevant to the article (so no sister ships of the same design, no stadiums where a game took place on a different day, and so on), it goes in if at all possible. I then rearrange the lists to try to hve enough text between each image that they don't start pushing down the one below them, with taller images (as they take up more vertical space) ideally getting cut first if there's too many. Still means that "good quality image" might have some bias, but it's about as fair as I can manage. I don't love the idea of having to choose which articles and images to highlight as well, which puts in all sorts of extra bias.

Don't get me wrong, I'm not saying any of the articles are bad or anything. But Wikipedia isn't a group that came together to share interests I selected, we came to try to cover the sum of human knowledge. The things that interest me most aren't going to be the things that interest you most, and I'm writing for you, dear reader, not for me.

This one is, unfortunately, quite short, because your Signpost writer is dealing with quite a few personal issues connected to his country locking down yet again. Maybe we should just cancel 2021. The last time I remember being unambiguously happy was back in February 2020, when I went to the Kirkcaldy Gilbert and Sullivan Society brilliant performance of The Gondoliers which, although I didn't know it at the time, would be the last live theatre I'd get to go to before everything shut down, cancelling a whole host of plans. The theatres shut down so abruptly that the Edinburgh Gilbert and Sullivan Society had brought in the sets and begun the tech rehearsals for Patience on stage, which would have had a performance had the theatres stayed open one more day.Edinburgh is a major theatre town - it hosts the Edinburgh Festival, has at least eight major theatres, and a lot more if you're willing to travel half an hour by train, and it's felt like the city's been dead for some time now.

-Adam.