Wikipedia:Wikipedia Signpost/Single/2019-04-30

Additional contributors: Pythoncoder

• As part of Wikipedia's April Fools' Day festivities, a recently created Wikipedia newspaper called TheWikiWizard was nominated for deletion at MfD. Signpost writer DannyS712 also writes for it.
• This year there were 2 successful RfAs: Linguist111 (testing) (closed at <10/9/1>), and DeprecatedFixerBot (closed at <7/13/6>). There was not a single unsuccessful RfA filed on April Fools' Day. RfBs, however, didn't fare as well: while ZLEA2 was closed as successful (with a final tally of <?/?/?>), ClueBot NG was unsuccessful, with their RfB closing at <17/16/12>.
• On the other hand, no humor BRFAs were denied: DeletionBoot was approved for trial, while C-3PO is still pending.

– D & P

The Signpost is not the only active publication here on the English Wikipedia. Other regular distributions include the Administrators' newsletter, the Guild of Copy Editors' reports, the Scripts++ newsletter, and WikiProject Portals' regular updates. Global publications are also available for reading, including Meta-Wiki's Tech News, Wikidata's Status updates, or the French Wikipedia's Wikimag. A new directory includes inactive newsletters and syndications of the Signpost. Editors are encouraged to add missing newsletters to the directory.

– D

• A new global special page, meta:Special:UrlShortener, was introduced.
• ArbCom accepted its first case since February.
• Three administrators were desysopped due to inactivity in April.

– D

Twenty-four featured articles were promoted this period.

• 
  Echo parakeet
• 
  French battleship Bretagne
• 
  Compulsory figures
• 
  27th Infantry Division Savska
• 
  Jaekelopterus
• 
  Lorde
• 
  Japanese aircraft carrier Hiyō
• 
  Norfolk, Virginia, Bicentennial half dollar
• 
  Marchioness disaster
• 
  Salih ibn Mirdas
• 
  Worcestershire v Somerset, 1979
• 
  The Colossus of Rhodes (Dalí)
• Hurricane Connie
• 
  Michael Collins (astronaut)
• 
  Tropical Depression Nineteen-E (2018)
• 
  Al-Mu'tadid
• 
  Lancaster's chevauchée of 1346
• 
  Irakli Tsereteli
• 
  Newberry Volcano
• 
  Crater (constellation)
• 
  Kim Clijsters
• 
  Round the Horne
• 
  SMS Lothringen
• 
  Western Area Command (RAAF)

Ten featured lists were promoted this period.

• 
  List of Casualty specials
• 
  Mac Miller discography
• 
  List of Cardiff City F.C. players (25–99 appearances)
• 
  List of Relic Entertainment games
• 
  List of accolades received by Mersal
• 
  List of awards and nominations received by Kylie Minogue
• 
  List of Billboard 200 number-one albums of 2001
• 
  List of German airplay number-one songs of 2018
• 
  List of Most Played Juke Box Folk Records number ones of 1944
• 
  Newbery Medal

Thirty-five featured pictures were promoted this period.

• 
  Anguish by August Friedrich Schenck
• 
  Stephen Hawking during a zero-gravity flight
• 
  Bar-tailed godwit (Limosa lapponica)
• 
  Double-banded plover (Charadrius bicinctus)
• 
  White-throated kingfisher (Halcyon smyrnensis)
• 
  Thérèse poster
• 
  Spotted wood owl
• 
  Pacific reef heron (Egretta sacra)
• 
  Plumed whistling duck (Dendrocygna eytoni)
• 
  Cardiac cycle
• 
  Ruins of Rudra Mahalaya Temple
• 
  Paulette del Baye
• 
  Marie-Aimée Roger-Miclos
• 
  Louis Désiré Blanquart-Evrard
• 
  Aftermath of the 1906 San Francisco earthquake
• 
  Dallol
• 
  Carmen poster
• 
  Westerhout 40
• 
  Mating common blue butterflies (Polyommatus icarus)
• 
  Bank myna (Acridotheres ginginianus)
• 
  Umberto Eco
• 
  Lucy Arbell
• 
  Orange-headed thrush (Zoothera citrina innotata)
• 
  Crested shriketit (Falcunculus frontatus)
• 
  Nelly Diener
• 
  Julia Margaret Cameron
• 
  Célestine Galli-Marié
• 
  Adonis blue butterfly (Polyommatus bellargus)
• 
  Bird-cherry ermine (Yponomeuta evonymella) caterpillars
• 
  Violet-backed starling (Cinnyricinclus leucogaster verreauxi)
• 
  Michael MacNamara
• 
  101955 Bennu
• 
  Black hole of Messier 87
• 
  Bronze grass skink (Eutropis macularia)

1700 map of Brest:

The Arbitration Committee resolved by motion to amend the "Conduct of arbitrators" section of the Arbitration Policy. The entire motion is reprinted below.

Pursuant to the policies for changing the arbitration policy (Wikipedia:Arbitration/Policy#Ratification and amendment), the community had to decide whether to amend the Arbitration policy or not. Specifically, the proposed amendment had to undergo a community referendum, and would only enter into force once it receive[d] majority support, with at least one hundred editors voting in favour of adopting it. That referendum was held at Wikipedia:Arbitration/Policy/Proposed amendment (April 2019), and was closed in favor of the proposed amendment.

A case request was submitted by Robert McClenon on 23 March 2019. The request attracted statements from 28 users, amassing more than 100,000 bytes of content. In the end, however, the committee declined the case request on 4 April 2019, with 2 Arbitrators voting to accept the case while 6 preferred declining it. See the archived discussion here.

As highlighted in this issue's news and notes, this month saw the successful RfA of RexxS following consensus at a bureaucrat chat. However, Pudeo asked the Arbitration Committee to review the 'crat chat, arguing that The bureaucrats' actions are not within the mandate they have been given and requesting that the Committee consider 1) affirming the 2015 election reform RfC, 2) review whether Maxim and possibly other bureaucrats acted within policy, and 3) overturning the bureaucrat chat. The Committee voted 10-0 to decline the case. The full case request can be found here.

After a hiatus that lasted over 2 months, the Arbitration Committee opened its second case of the year on 13 April 2019. As of writing, the scope of the case is limited to examining The administrative conduct of Enigmaman [and] What action is required, if any. See the prior (archived) discussion at ANI here, and the full case itself here.

Last month, Necrothesp's administrative permissions were removed under level 1 procedures "as a suspected compromised account". This month, the committee resolved to return their bit. Furthermore, the Arbitration Committee has taken steps to encourage administrators to improve the security of their accounts.

The Arbitration Committee resolved by motion to amend the "Return of [administrator] permissions" section of the Arbitration Procedure. The entire motion is reprinted below.

Back in February, GoldenRing deleted User:Dlthewave/Whitewashing of firearms articles, claiming that this was an Arbitration Enforcement action. The deletion was brought up at Deletion review, during the course of which Bishonen restored the page so that it could be viewed by non-administrators. At issue was whether deletion is a valid Arbitration Enforcement action, and, separately, whether it was proper to consider overturning an action labeled to be enforcing an arbitration decision at deletion review. To clarify proper procedures in such cases, the Arbitration Committee resolved by motion to amend the Standard provision for appeals and modifications in the Arbitration Procedure. The entire motion is reprinted below.

• By motion, the Arbitration Committee lifted SheriffIsInTown's topic ban from pages related to conflict between India and Pakistan, subject to a six month probation period
• The Arbitration Committee is getting a new bot to assist its clerks. ArbClerkBot, operated by Bradv, is currently in trial.
• The Committee is discussing two requests for clarification and amendment: one relating to Eastern Europe, and the other dealing with Palestine–Israel articles.

This traffic report is adapted from the Top 25 Report, prepared with commentary by Igordebraga (March 24 to April 13), Pythoncoder (April 6 to 13) and Stormy clouds (April 14 to 20).

Rank	Article	Class	Views	Image	About
1	Notre-Dame de Paris		6,716,199		One of the more recognisable Parisian landmarks, alongside Le Tour Eiffel and L'Arc de Triomphe, was irrevocably changed by a monumental blaze this week, which captivated and shocked the world. Renovation works were underway on the famed cathedral, and these reportedly exacerbated the risk of, and ultimately caused, the fire, which consumed the spire of the church, alongside much of the roof. Firefighters nobly attempted to contain the inferno, but given the impossibility of audacious plans involving airdropping water, were unable to prevent much of the damage sustained to the historic 12th century site, whose striking Gothic architecture may never be the same again. Thankfully, the tragic damage sparked a wave of generosity that culminated in charitable donations amounting to €900 million to finance the subsequent renovations and repairs, which may take up to two decades to be completed.
2	Tiger Woods		3,127,942		The greatest sporting comeback outside of Constantinople took place on the fabled fairways and treasured tees of Augusta National last weekend, with the long-awaited return of the most successful golfer of the 21st century to the most major of stages. For the first time in eleven years, the world of professional golf heard the roar of the ferocious tiger, mercilessly efficient in pursuing his prey through Amen Corner and beyond, and ultimately donning the green jacket for the fifth time in his storied career, a feat which would have been considered an impossibility by the vast majority of golf aficionados a mere year ago.
3	Game of Thrones (season 8)		2,528,460		The best show currently on television has returned for its eighth and final season, and the world is on tenterhooks to see who shall sit upon the Iron Throne once the dead have been defeated and the dust has settled, as evidenced by the gargantuan number of page views that articles associated with the HBO series have seen in recent weeks. Given that Game of Thrones is renowned for its shocking, horrific, traumatising deaths, you should all be prepared for the only logical way in which the series can conclude - with the commencement of the reign of Hot Pie, first of his name.
4	Game of Thrones		1,516,505
5	Billie Eilish		911,973		I honestly have very minimal idea who this is, beyond the fact that she is a singer, and after listening to some of her music in preparation for this piece, have minimal inclination to change this. My exposure to Eilish involves hearing various jokes about how her new single resembles the greatest song ever written, but, to reiterate, I lack the insight to either verify or refute this claim. Only time will tell if the emergent seventeen year-old is a new force in the music industry, or a flavour of the month, and I am hesitant to speculate either way. To do so would surely make me the bad guy.
6	List of Game of Thrones episodes		900,781		Duh duh duh na nuh nuh da da duh duh duh na nuh nuh ...
7	Notre-Dame de Paris fire		863,794		The fire which occurred at the gorgeous Gothic cathedral in Île-de-France caused extraordinary destruction, but more positively united the people of France and beyond in charitable unity, spearheaded by nine-figure donations from famous French billionaires like the Arnault, Bettencourt, and Pinault families, which will be used to finance the necessary reconstruction and restore the cathedral to its world-famous, awe-inspiring glory.
8	Pete Buttigieg		760,812		The mayor of South Bend, Indiana, Pete Buttigieg is the latest of a long string of donkeys to throw their hats into the ring to contest the 2020 election against the Donald. Buttigieg lacks the credentials of some of his forthcoming opponents, having never served in either chamber of Congress, but the Harvard graduate has garnered a considerable amount of attention from the media after embracing his husband on stage at his candidacy announcement, propelling him to unprecedented heights in primary polls.
9	Bonnie and Clyde		728,453		The most notorious amourous duo of American crime are currently capturing the eyes of binge-streaming Wikipedians courtesy of the content overlords over at Netflix, and their latest viral original film, The Highwaymen, which stars the Prince of Thieves and Haymitch Abernathy as FBI agents in pursuit of the murderous thieves, who have somehow been romanticised in spite of their status as serial killers. The couple have maintained a constant presence in the pantheon of criminals for nearly a century, aided by zeitgeist-capturing appearances such as those of Dunaway and Beatty, and thus it is of little surprise that they continue to engage the curiosity of Wikipedians.
10	Avengers: Endgame		726,105		I lack the courage to mutter a mere word, or even a Google search, until midnight on Thursday, so sorry. You know what this is, and should ready yourself for it to rest atop future iterations of the report - this article's page views are about to tend to infinity.

• These lists exclude the Wikipedia main page, non-article pages (such as redlinks), and anomalous entries (such as DDoS attacks or likely automated views). Since mobile view data became available to the Report in October 2014, we exclude articles that have almost no mobile views (5–6% or less) or almost all mobile views (94–95% or more) because they are very likely to be automated views based on our experience and research of the issue. Please feel free to discuss any removal on the Top 25 Report talk page if you wish.

• Notifier^[1] (source) – quickly notify other talk pages of a discussion on the current page
• View Source^[2] (source) – Adds a "view source" option to editable pages
• Restorer^[3] (source) – adds links beside every revision in page history to restore that revision.

• JJMC89 bot III (BRFA · contribs · actions log · block log · flag log · user rights) - an ADMINBOT
• RetractionBot (BRFA · contribs · actions log · block log · flag log · user rights)
• FastilyBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 14)
• KadaneBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 3)
• RF1 Bot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 2)
• GreenC bot (BRFA · contribs · actions log · block log · flag log · user rights) (Tasks: 13, 14)
• PkbwcgsBot (BRFA · contribs · actions log · block log · flag log · user rights) (Tasks: 7, 12, 23)
• DannyS712 bot (BRFA · contribs · actions log · block log · flag log · user rights) (Tasks: 12, 15, 16, 17, 23, 25, 27, 28, 29, 30, 31)

Open

• DiBabelYurikBot (BRFA · contribs · actions log · block log · flag log · user rights)
• SportsStatsBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 2)
• JATMBot (BRFA · contribs · actions log · block log · flag log · user rights)
• Xinbenlv bot (BRFA · contribs · actions log · block log · flag log · user rights)

In trial/extended trial

• GreenC bot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 11)
• ArbClerkBot (BRFA · contribs · actions log · block log · flag log · user rights)
• WikiCleanerBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 2)
• HostBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 9)
• PkbwcgsBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 5)
• DannyS712 bot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 13)

Trial complete

• DannyS712 bot (BRFA · contribs · actions log · block log · flag log · user rights) (Tasks: 33 & 34)
• Qbugbot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 3)
• WikiCleanerBot (BRFA · contribs · actions log · block log · flag log · user rights) (Task: 3)

Latest tech news from the Wikimedia technical community: 2019 #14, #15, #16, #17, & #18. Please tell other users about these changes. Not all changes will affect you. Translations are available on Meta.

New and upcoming features

• When you add an edit summary the VisualEditor will search your recent edit summaries in case you want to re-use one. This works in both the visual and wikitext modes on desktop. It also works on the mobile site. [1]
• You will be able to preview references. This means that when you hover over the link you will get a popup that shows you a preview of the reference. It will work much like page previews. This is so you don't have to go to the bottom of the page to see a reference. This will now be available as a beta feature on German and Arabic Wikipedia. [2][3]
• The advanced search function URL now shows which namespaces you search in. The namespace field is collapsed by default on the search page. You can also add new fields to the search interface through a hook. [4][5][6]
• The wikis now look slightly different in the mobile web version. [7]
• Wikipedia articles will have the sameAs meta property. It adds structured data. This makes it easier for search engines to find Wikipedia articles. It also makes it easier to reuse content. There will an A/B test. [8][9]
• You will see when you last refreshed the recent changes page. This is so you can see how recent the changes are. [10]
• When you write a comment in Structured Discussions but have not posted it yet your web browser will save it in local storage instead of session storage. This means you do not lose them even if you close your web browser. Structured Discussions used to be called Flow. [11]
• You will be able to turn off milestone notifications. Milestone notifications congratulate you when you have made certain numbers of edits. [12]
• Wikimedia wikis will soon use a token when you log out. This changes how the API works. Some tools might need to be updated. [13]

Old features

• Notifications tell you about things that happen on the wiki. You can turn on notifications about new links to pages you created. For performance reasons you can no longer get e-mails about this. [14]
• The {{REVISIONID}} magic word will no longer work. This is for performance reasons. When you preview a page it will return "" (empty string). When you read a page it will return "-" (dash). For now this will only affect content namespaces. [15]

Problems

• A map update caused some problems on 29 March and 30 March. It was rolled back. [16]
• Pages on some Wikivoyages had problems with the top headline. This has been fixed. [17]
• Tech News was late last week because of a MassMessage bug. Other newsletters had the same problem. [18]

Wikidata

• Wikidata will get a new constraint status called suggestion. This will change how the WikibaseQualityConstraints constraint checking API works. [19][20]
• The Wikidata JSON output will change. Empty containers will be serialised as empty objects. This is a breaking change that will affect tools that use JSON outputs and APIs. It will happen on 30 April. You can read more and see how to test your code.
• The Wikidata wb_terms table will be dropped. This will affect some Wikidata tools. They need to be updated. The table has become too big which is causing problems. This will happen on 29 May. You can read more. You can ask for help if you need it.

Other updates

• All wikis now have the TemplateWizard for the wikitext editor.
• You can test the depicts property for structured data on Commons.
• More wikis are now testing visual editor section editing for mobile users. You can read more.
• Admins on French and Polish Wikipedia can block someone from editing a page or a namespace. If your wiki wants to get these blocks before they are available to all wikis, you can tell the developers. [21][22]
• The Wikipedia app for Android now invites users to add Wikidata descriptions to Wikidata objects that have Wikipedia articles but no Wikidata descriptions. It will only invite users who have added a number of Wikidata descriptions in the app without being reverted. This is to avoid spam and bad edits. [23][24]

• You can join the technical advice meeting on IRC. During the meeting, volunteer developers can ask for advice. The meeting takes place every Wednesday from 4:00–5:00 p.m. UTC. See how to join here.

This article originally appeared in the Wikimedia Foundation blog on April 3, 2019.

We are using machine learning to predict whether—and why—any given sentence on Wikipedia may need a citation in order to help editors identify areas of content violating the verifiability policy.

One of the key mechanisms that allows Wikipedia to maintain its high quality is the use of inline citations. Through citations, readers and editors make sure that information in an article accurately reflects its source. As Wikipedia’s verifiability policy mandates, “material challenged or likely to be challenged, and all quotations, must be attributed to a reliable, published source”, and unsourced material should be removed or challenged with a citation needed flag.

However, deciding which sentences need citations may not be a trivial task. On the one hand, editors are urged to avoid adding citations for information that is obvious or common knowledge—like the fact that the sky is blue. On the other hand, sometimes the sky doesn’t actually appear blue—so perhaps we need a citation for that after all?

Scale up this problem to the size of an entire encyclopedia, and it may become intractable. Wikipedia editors’ time is limited and their expertise is valuable—which kinds of facts, articles, and topics should they focus their citation efforts on? Also, recent estimates show that a substantial proportion of articles have only a few references, and that one out of four articles in English Wikipedia does not have any references at all. This suggests that while around 350,000 articles contain one or more citation needed flags, we are probably missing many more.

We recently designed a framework to help editors identify and prioritize which sentences need citations in Wikipedia. Through a large study that we conducted with editors from English, Italian and French Wikipedia, we first identified a set of common reasons why individual sentences in Wikipedia articles require citations. We then used the results of this study to train a machine learning model classifier that can predict whether or not any given sentence needs a citation —and why—on the English Wikipedia. It will be deployed in the next 3 months to other language editions.

By improving the identification of where Wikipedia gets its information from, we can support the development of systems to help volunteer-driven verification and fact-checking, potentially increasing Wikipedia’s long-term reliability and making it more robust against biases, information quality gaps and coordinated disinformation campaigns

Why do we cite?

To teach machines how to recognize unverified statements, we first needed to systematically classify the reasons why sentences need citations.

We started by examining policies and guidelines related to verifiability in the English, French, and Italian Wikipedias and attempted to characterize the criteria for adding (or not adding) a citation described in those policies. To verify and enrich this set of best practices, we asked 36 Wikipedia editors from all three language communities to participate in a pilot experiment. Using WikiLabels, we collected editors’ feedback on sentences from Wikipedia articles: editors were asked to decide whether a sentence needed a citation and to specify a reason for their choices in a free-text form.

Our methods and our final set of reasons for adding or not adding a citation can be found on our project page.

Teaching a machine to discover citation gaps.

Next, we trained a machine learning model to discover sentences needing citations, and characterize them with a matching reason.

We first trained a model to learn from the wisdom of the whole editor community how to identify sentences that need to be cited. We created a dataset of English Wikipedia’s “featured” articles, the encyclopedia’s designation for articles that are of the highest quality—and also the most well-sourced with citations. Sentences from featured articles that contain an inline citation are considered as positives, and sentences without an inline citation are considered as negatives. With this data, we trained a Recurrent Neural Network that can predict whether the sentence is positive (should have a citation), or negative (should not have a citation) based on the sequence of words in the sentence. The resulting model can correctly classify sentences in need of citation with an accuracy of up to 90%.

Explaining algorithmic predictions

But why is the model up to 90% accurate? What is the algorithm looking at when deciding whether a sentence needs a citation?

To help interpret these results, we took a sample of sentences needing citations for different reasons, and highlighted words the model considered the most when it classified the sentences. In the case of “opinion” statements, for example, the model assigned the highest weight to the word “claimed”. In the “statistics” citation reason, the most important words to the model are verbs that are often used in reporting numbers. In the case of scientific citation reasons, the model pays more attention to domain-specific words like “quantum”.

Predicting why a sentence needs a citation

Similar to the “reason” field of the [citation needed] tag, we want our model to also provide full explanations of citation reasons. Therefore we created a model that can classify statements needing citations with a reason. We first designed a crowdsourcing experiment using Amazon Mechanical Turk to collect labels about citation reasons. We randomly sampled 4,000 sentences that contain citations from Featured articles, and asked crowdworkers to label them with one of the eight citation reason categories we identified in our previous study. We found that sentences more likely need citations when they are related to scientific or historical facts, or when they reflect direct/indirect quotations.

We modified the neural network designed in the previous study, so that it can classify an unsourced sentence into one of the 8 citation reason categories. We retrained this network using the crowdsourced labeled data, and found that it provides reasonable accuracy (precision at 0.62) in predicting citation reasons, especially for classes with a substantial amount of training data.

Next steps: predicting “citation need” across languages and topics

The next phase of this project will involve modifying our models so that they can be trained for any language available in Wikipedia. We will use these multilingual models to quantify the proportion of unverified content across Wikipedia editions, and map citation coverage across different article topics, in order to help editors identify areas where adding high quality citations is particularly important.

We plan to make the source code of these new models available soon. In the meantime, you can check out the research paper, recently accepted at The Web Conference 2019, its supplementary material with detailed analysis of the citation policies, and all the data we used to train the models.

We would love to hear your feedback and comments, so please reach out to us on our project page to help us improve it.

The authors would like to thank the community members of the English, French, and Italian Wikipedias, along with workers from Amazon Mechanical Turk, for helping with data labeling and for their precious suggestions.

Miriam Redi is Research Scientist at the Wikimedia Foundation
Jonathan Morgan is Senior Design Researcher at the Wikimedia Foundation
Dario Taraborelli is a former Director of Research at the Wikimedia Foundation
Besnik Fetahu is a Post-doctoral Scientist at the L3S Lab Hannover

The future of portals

In June Katherine Maher will mark her third anniversary as executive director of the Wikimedia Foundation (WMF). Before becoming ED, Maher served as acting ED for three months, and WMF chief communications officer for two years. She had previously worked in communications with UNICEF, the National Democratic Institute, the World Bank, and Access Now.

This interview covers topics from the WMF's relationship with the editing community, to her accomplishments as ED, to diversity, harassment, WMF spending, airplanes, global advocacy, and The Badass Librarians of Timbuktu.

Sign Post: When you became ED what were your feelings then about the tasks ahead of you and in particular about working with the editor community which was then in an uproar? Is working with the community of editors one of your most challenging tasks?

Katherine Maher: Well, it was a really unexpected new role, and I was focused on trying to live up to the needs of the movement. I’d also just had back surgery for a spinal injury four weeks before Wikimania Esino Lario, so I was literally shaky on my feet! Which is to say, my feeling were a real jumble: I was humbled by the faith from the Board, I was excited for the chance to set the Foundation on a course that was more aligned with my understanding of our movement than in the past, and I was anxious about taking seriously the importance of the work ahead.

Truthfully, the relationship between the Foundation and the community had been strained for as long as I’d been around. There was little trust between the two groups, and the Foundation had made plenty of missteps along the way. But suddenly, Foundation staff, who are truly dedicated to the movement and mission, found themselves also in a difficult position, and more aligned with community than at any time in the past. There was a lot to take on at once, but also a lot of opportunity to try and reset and realign folks around the bigger free knowledge vision and our shared sense of values.

I felt as though there were three truly critical things to do: Improve the relationship between the Foundation and the community (both the individual editor community and also our community of affiliate groups), support Foundation staff coming out of a period of cultural upheaval, departures, and transitions, address the issues and give people a sense of stability, and rebuild critical capacities and institutional knowledge.

Bring people together around a direction for the organization and the movement overall -- a shared purpose and set of values. This became our work on movement strategy and creating a shared sense of direction across the movement.

To me, all three of these things were urgent and important.

You asked specifically about working with the editor community. I believed the tension between the editor community and the Wikimedia Foundation was untenable for community and staff alike. We’re all here for the vision, and being at odds with one another was going to hurt this thing that we’ve spent so many years building. I knew the Foundation needed to spend more time listening to communities, understanding people’s challenges and frustrations, and taking steps to respond. I didn’t expect that we would agree on everything - our movement is far too complex and diverse for that. But my goal was -- and continues to be -- to build trust and move forward. Today, I think we’re in a place where we're having more of a conversation, even if we don’t agree with each other. That’s a far better place!

So, in general, I feel like working with the editor community isn’t about being hard or easy -- it’s part of who we are, and how we make Wikimedia possible. When it’s hard, we at the Foundation have to ask ourselves why: is it because we’re off track? Because we need more conversations? Because we’ve skipped a step of trust? Again, sometimes we’re going to disagree, but let’s try to do that openly, with good faith. Because sometimes we’ll change our minds, sometimes the community sentiment will change, and sometimes we’ll find the right compromise. Such a big part of the Wikimedia ethos is that none of us necessarily knows more than the others of us, so there’s always an opportunity to learn together on how we can build and sustain this incredible thing on behalf of so many.

SP: What have been the major accomplishments of Wikipedia and the WMF in the last three years? What have been your own major accomplishments in the last three years?

KM: There are so many things that the movement has accomplished over the past three years. We’ve celebrated the 18th birthday of the English Wikipedia, and milestone birthdays for many other projects. We've seen our 300th language's Wikipedia emerge. Commons and Wikidata are up to 50 million images and items, respectively. Our community keeps growing globally, with more than 100 global affiliates, and women are increasingly represented in leadership roles. I like to point to the fact that on English Wikipedia, the percentage of articles about women has increased from 15 to nearly 18 percent. 15 to 18 percent may not sound like much, but that represents more than 86,000 new articles, about 72 new articles about women every single day, for about three and a half years -- and usually these articles are more meticulously sourced and higher quality than your average article. The ability to move the needle one article at a time is significant.

In the context of the broader world, I’m also excited that Wikipedia is more trusted today than ever. I don’t mean in the sense of mindless consumption of the articles -- after all, we’re a “citation needed” culture. Instead, I mean that more people see Wikipedia as a place that they can come to learn, to discuss, and find a sense of shared understanding -- that it’s precisely because of our non-commercial, community-driven model that they can trust us. At a time in which so much information is now perceived as contentious or polarized, Wikipedia’s importance to the public discourse is something that we can be proud of. The community and the Foundation share something of critical importance: we’re all working to improve Wikipedia. In that sense, the achievements of the Foundation are usually achievements shared with the community.

I'm proud of working with so many different community members on the strategic direction process. Through that process we’ve come together to identify our challenges, and collaborated to build something that is consistent with our movement, yet also pushes us to take on a broader, more inclusive understanding of what knowledge is and who gets to participate in knowledge. I'm also excited by the investments that we've been making in Wikimedia Commons and in Wikidata, as the Foundation is beginning to expand its perspective to their importance in the Wikimedia ecosystem. Finally, I think it’s important that we’ve been developing a voice that stands for our values in policy discussions and in broader public discourse.

Another way to think about this: when I came in as the head of communications several years ago, I used to open up the press clippings every day, and read with dismay articles about how Wikipedia was dying. Today, people are talking about how unique, important, and essential Wikipedia is. I think that is perhaps one of the most significant things that we've all done together.

SP: Both you and the WMF are unapologetically pro-diversity and more specifically, feminist. The number of women editors has not risen to 25% which was a goal set several years ago. How have you encouraged editors working with Women in Red and women’s issues in general over the last 3 years? What more do you plan to do?

KM: First of all, I take that as a compliment. 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. Our obligation to all the world's knowledge is about feminism and gender identity, along with other forms of diversity. Having said that, I share our critics’ dismay around our challenges in moving the needle on gender representation in our editing community and in our content. And when I speak about these issues, I'm rarely the first person to raise the alarm -- if anything, I’m often echoing back things that community leaders have been saying for some time.

It’s true that we haven't reached the 25% overall participation rate for contributors to our projects, but we’re seeing positive movement in other parts of the movement. According to a 2016-2017 community survey, women represent 28% of affiliate leaders and 25% of program organizers, and the percentage of female representatives in affiliates’ boards grew from 20% to 28% between 2013 and 2017 . We also have much better gender representation when you look at chapter EDs, Board seats, affiliate committees and the like.

At the Foundation we’ve been focused on providing support, training, and mentoring for groups that focused on these issues. This runs the gamut, from researching and highlighting their work on our blog, such as Emily Temple-Wood and the “Keilana effect,” supporting peer-learning gatherings such as Wiki Women Camp and 2018’s Diversity Conference, or direct funding for intersectional efforts such as Art + Feminism. The core organizing team of Art+Feminism has received six grants over the past five years from the Foundation, totaling more than $440,000, as well as 32 grants to local organizers in many countries, including Egypt, Ghana, India, Peru, Botswana, New Zealand, Cameroon, Japan, and Italy. For many of these communities these are the first events organized around gender and have catalyzed both conversations and additional initiatives to increase gender diversity on the Wikimedia projects.

I also think it matters that the Wikimedia Foundation is a majority female-led organization. We have a woman board chair, a woman Executive Director, and more than 50% of people in management and leadership positions at the Wikimedia Foundation identify as women. I genuinely believe that this matters -- not just as a model, but in the decision-making of the organizations, affiliates, and structures of power that make up our movement. I’ll be curious how the impact of this leadership plays out in five years time.

All this said, we know we have a ways to go. There's work that we can do from our product standpoint, and in terms of supporting programs. The Foundation is currently doing a global survey* to understand women's perceptions around the value they get from Wikipedia and how we can encourage more women to participate. Our initial findings support what many in our movement have presumed anecdotally about how women engage with Wikipedia and why. Globally, more men have tried to edit Wikipedia than women have, and men find it more useful and trustworthy than women do. We’ll be sharing more information about these results and how our learnings will inform our future work to bring more women into our movement the coming months.

*The survey was run in a representative sample from the United States, Nigeria, Germany, Egypt, India, and Mexico.

SP: The New York Times recently wrote about harassment on Wikipedia targeting transgender and women editors. Have Wikipedia and WMF policies been effective in preventing harassment? Would the WMF or administrators offend the editing community as a whole by strict enforcement of anti-harassment rules, or by instituting new rules? The Times article stated that the WMF is considering a new private harassment reporting system. How is that going to work and when will it be rolled out?

KM: We know that harassment on Wikimedia is real. And it can be even more challenging when compared to other major web platforms, because harassment in Wikimedia is rarely as straightforward as that in the comments section of a social news feed. It’s just as problematic, but often harder to detect using technical tools. For example, when we started looking into harassment on the English Wikipedia, we learned that starting a conversation with the word “Please” was a strong indicator that a conversation would go off the rails. You wouldn't think that “Please,” an otherwise polite word, would lead to an unpleasant interaction -- and yet, on Wikipedia it often does. In my mind, this underscores the differences we face in tackling these issues versus other organizations. the relationship between the Foundation and the community had been strained for as long as I’d been around. There was little trust between the two groups, and the Foundation had made plenty of missteps along the way.

You asked about Wikipedia’s policies around harassment. Truthfully, I don’t think most of the on-wiki policies we have today are conducive to creating safe and welcoming spaces. Far too frequently, people use our policies to walk the line while still engaging in harassing behavior. These are policies that we as a community have built together over time, and they enable behavior that drives away newbies, minorities, and other marginalized groups, making our mission unquestionably poorer for the absence of their voices. I believe that it’s time for the community and for the Foundation to engage in conversation about whether these policies are ultimately serving our objective of building the encyclopedia -- or whether they're preventing us from doing so.

In 2016 when I became Executive Director, the Board gave the Foundation a mandate to address harassment. Since then, we’ve become more assertive about enforcing actions on repeat offenders in our communities. In addition to removing the worst offenders, I hope people also feel secure in knowing that there are certain types of behavior that will no longer be tolerated. Similarly, the anti-harassment team at the Foundation has been researching and rolling out efforts around clear, enforceable policies and better tools to identify and address harassing behavior, but there’s still much to do.

One of the strategies we’re considering next is how do we improve the reporting process, both for the person flagging harmful behavior and for the community members who take action. We want to help people make more informative reports that get routed to the right person and that have a clear path to escalation if needed. We’re also looking at how we continue to improve transparency around these processes, to help both the community and the Foundation better understand the volume and severity of issues.

Yet, while that’s a good start we know there’s more work to be done. I often think about how great community members are at recognizing and addressing copyright violations on our projects -- imagine if our community felt that comfortable and empowered addressing harassment?

SP: Since 2017 when the WMF anti-discrimination policy was revised, there are no WMF policies against discrimination of protected classes of editors, only against discrimination against WMF employees and contractors. There would be some obvious difficulties in protecting anonymous editors against discrimination and it would be a huge job in any case, but is the WMF considering doing anything in this area?

KM: In 2017, we updated the policy to include new explicit protections and expanded definitions related to gender identity and expression, disability, citizenship, and ancestry. Historically the policy had been applied to only staff and contractors, and we wanted to make its application more explicit. Our intent wasn’t to limit how this was applied to people contributing to the projects, but to create opportunity for volunteers to explore spaces for women-identifying groups or cultural affinity groups on-wiki which might otherwise have been limited by the language of the previous policy. Of course, we still have other guidance on engagement that applies, including resources such as the code of conduct for technical spaces.

(As I was answering this question, I realized that I actually wrote the 2017 announcement about our changes to the anti-discrimination policy!)

SP: There’s a very old criticism of the WMF that predates your employment at WMF. It keeps on resurfacing from time to time however. How do you react to editors who say “The WMF has grown too fast; it raises and spends too much money. Much of that spending is just wasted on too many employees who don’t directly help editors.”?

KM: I know this one! And I’m sympathetic. We’re first and foremost a volunteer community and project, and there’s a natural tension with how resources are allocated between volunteer efforts and staff -- that’s true in so many organizations. Furthermore, how you see the Foundation’s size and budget is all a matter of perspective: after working for a 20-person non-profit, 350 people seems enormous to me. But compared to tens of thousands of people at other large web platforms, 350 often feels like a raft on the ocean.

Every year at the Foundation, we go through a budgeting process in which we look at everything we’re doing, and figure out where we might be able to hire a few new people. And every year, we see increased needs and demand -- whether it’s improved site performance and reliability, more support for non-Latin languages on-wiki or in Foundation communications and services, improved interfaces for mobile editing, improved tools against harassment, more support for diverse geographies, upgraded security and privacy practices, new tooling for the sister projects -- there’s no shortage of what we could be doing. And we want to do it all with independence and integrity, in a way that reflects our values. We might have fewer staff and a smaller budget if we supported fewer languages, used private cloud hosting, or didn’t prioritize community health, but is that the Wikimedia we want to be?

Another way to think about this is that Wikimedia is the fifth most popular website on the planet, with about a billion people globally using our sites. Everyone else even remotely near this scale has tens of thousands of employees. And yet, even with those numbers, we’re so far from realizing our true vision: a future where every single person in the world can share in the sum of all knowledge. Often, when I talk to Foundation staff, they tell me we could be 3,500 people and still not cover the scope of the mission or the community needs. If we only reach around one billion people right now, we’re definitely going to need more capacity to realize that aspirational goal.

But that’s my perspective. And as I said at the beginning of this interview, we’re not always going to agree and that’s okay. That’s part of the nature of how Wikimedia works. I recognize how fortunate our movement is to have the amount of financial and community support that we do -- we should approach it with respect and humility.

SP: You spend a great amount of time on airplanes. Could you give an estimate of the number of flights or number of miles you’ve flown in the last year? Does this limit the amount of work you can get done in San Francisco? Who runs things while you are away? What are the benefits of the trips?

KM: Indeed -- it’s a lot of time on planes, too much, if you ask me! Last year, I think traveled about 200,000 miles, which is probably unhealthy for me and the world. But at the same time, I’ve found it absolutely invaluable to understanding the needs, challenges, and opportunities our communities face.

In any given month, I’m often attending community events and gatherings, including regional and thematic conferences and the Wikimedia Summit and Wikimania. Increasingly, I’m also on the road to meet with supporters of the movement -- not just donors, but partners and other allied groups. Sometimes I’m out speaking to critics: as pressure on “big tech” increases, I’m meeting journalists and policymakers to explain why our model and community is so different than many other parts of the web, and to encourage them to support policy objectives that recognize those differences.

I believe the biggest benefit of my trips is that I get to meet people in the places where they work, whatever that is. Wikimedia’s mission is most powerful when it comes to you: when you experience it in your context, your language, or as aligned with your personal mission. And meeting community members in their local environments is incomparable -- whether that’s at a library for an edit-a-thon, a university outreach program, or a gathering of a local affiliate. It is an opportunity to understand local challenges, put faces to names, hear about the impact of Wikimedia in a language or community, and truly understand the breadth and importance of our movement to so many people.

As for how it all works -- well, I’m answering these questions right now on a plane from 30,000 feet in the air! My travel doesn’t really limit my work, since the Foundation is increasingly remote and distributed across the world -- today only about a third of the Foundation’s 350 staff are in San Francisco. I strongly believe that this is the best way for our organization to reflect and serve our global community, which is itself remote and distributed, working collaboratively over distance and time. There's something very meaningful about working the way that the majority of our staff and contributors work. And fortunately, we have a great team and a new COO to make sure everything keeps moving in San Francisco -- and they know how to reach me, any time of day or night.

SP: Do you have time while flying to do much non-WMF reading? I was surprised that the Financial Times had two photos of your bookshelves, showing titles in Arabic, and authors like Daniel Kahneman and Philip Howard. Do these books contribute to your work at WMF or are they just to satisfy your personal search for knowledge? What are you reading now?

KM: Right now I’ve got three books in my carry-on bag: The Badass Librarians of Timbuktu, The Sympathizer, and Why I’m No Longer Talking to White People About Race. As I mentioned, I spend so much time on planes, and while it’s a great time to get work done, it’s also an opportunity to take a deep breath and learn about the world.

I was always a voracious reader as a little kid -- I knew my hometown librarians by name, and they all knew me. My family would visit the library with a canvas bag and fill it up with books -- and then a week later, we’d go back and swap them all out for more. I’ve kept this up as an adult, though I own more titles now. In fact, the books on my bookshelf are just a subset of the total -- my parents really wish I’d come collect the rest from their house!

The reason you've noticed titles in Arabic, German, and other languages on my bookshelf in that photo is that I did study in all of those languages at one point, although unfortunately none really stuck. As for Kahneman and Howard -- I’ve been fortunate to have known Professor Howard for some time, and learned a lot from his work! But generally, my bookshelf reflects my eclectic set of interests, the places and work I’ve spent time, and my abiding interest in how technology intersects with the best and worst of the human condition.

SP: One of the proposed goals in the 3-5 year plan is to “Modernize our product experience.” This sounds like a move away from websites to mobile phone apps or perhaps other technologies. How do you expect this goal to be accomplished?

KM: You’re absolutely right -- but I’d restate it less about “moving away” from websites and more about embracing a future in which the website experience is only a part of our digital ecosystem. Already, it’s predicted that by 2020, 50% of all search experiences will be through digital voice assistants. And that’s not just on fancy phones and home devices -- voice assisted technologies have the potential to be transformative in lower literacy communities. Wikipedia should always be a place for collaboration and creation - open source, read/write, and true to our values. But I believe the vision for Wikipedia and Wikimedia goes well beyond the websites, to a world of connected free knowledge -- one in which people can explore free knowledge on different devices, interfaces, and experiences, where they can ask simple or complex questions, immerse themselves in rich images and media, and connect to other free knowledge repositories beyond our ecosystem (as already happens through so many of our GLAM partnerships). I believe this means reaching people in parts of the world we’re not currently serving well by making contribution, curation, and collaboration tools more equitable and easy to use. That means providing tools that work across many different devices with minimal technical needs. Mobile will absolutely be a part of the solution, although it won’t be limited to just new mobile technologies. We can do a lot more with our existing AI and automation tools to find content gaps in our projects, protect the integrity of content, and empower smaller projects within the community to adapt our software and platform to what’s most effective for them.

To make all of this happen, the Foundation and movement will need to invest in machine learning, structured data development, multimedia and interactive content capacity, server and network infrastructure, developer tooling and engineering productivity, and volunteer diversity. We need to understand how people are experiencing and consuming knowledge globally, and adapt while staying true to our values.

SP: One of the proposed priorities in the 3-5 year plan is global advocacy. Any political advocacy is always controversial among editors. Ideally the editing community should include people of almost all political views and movements. How can you engage with governments as different as those in China, Iran, Mexico and the U.S. while not offending large parts of the editing community?

KM: Wikimedia is not and should never be a partisan political organization. But global advocacy is not about politics -- or not only about politics! It is about understanding the broader legal, regulatory, and policy context in which our movement operates -- what are the laws and policies that support us in better serving our mission, and what are the policy trends that could threaten our projects, platforms, and people? I believe we have an obligation to advance a policy agenda that supports free knowledge, as well as protect the open web and fundamental rights that enable our community, projects, and mission to thrive.

To manage the diversity of perspectives from the many different governments around the world, our conversations need to be grounded in our values and our overarching mission. Most of the time, our conversations with governments focus on education and awareness, explaining how Wikipedia works, while remaining rooted in our values and commitments. But sometimes, this will also mean engaging with difficult actors -- especially those who attempt to censor or block the projects. But as ever, we’ll never compromise on censorship, transparency, or the Wikimedia open model. Again, our communities may not always agree on which policies, which issues, how vocal, or when. But I’ve always been impressed by our ability to find and build on consensus, including in conversations about the policies and laws where Wikimedia’s perspective should be heard.

SP: Another proposed priority in the 3-5 year plan is to increase worldwide readership, especially in Asia and Africa. The WMF has always been aggressively international and multi-cultural in its work, yet readership is still concentrated in Europe and North America. What can you do differently to help broaden Wikipedia’s readership?

KM: We know that by 2030, 80% of the world’s population will be in Asia and Africa -- continents that currently rank lowest on Wikimedia participation and readership. If we judge the success of our mission by every single human participating in all the world’s knowledge, that means we really have a ways to go.

This is why increasing worldwide readership is a top priority -- creating a stronger global presence of Wikipedia readers requires intention and work across the movement, from raising awareness of Wikipedia in parts of the world where it’s not widely known, to conducting user research on how people search and discover knowledge on and off the projects, to evolving the way people find and access knowledge on Wikipedia -- whether that be through new formats like video or structured data.

It means increasing language and translation support, and supporting our communities so that when people come to Wikipedia for the first, second, or third time, they’re finding articles that are relevant to them, in their own language. This includes working with new and existing partners, including in education and in GLAM, who better understand local context and knowledge traditions. It’s much more than readership -- after all, readers won’t use the projects if they don’t see their language, culture, experiences reflected in them. Readership is just the biggest, hardest thing to take on -- so we want to set that ambitious goal, in order to better understand the equally ambitious actions we undertake to be successful.

SP: Everybody, at least occasionally, thinks about their next job. What type of work would it take to get you to consider leaving the WMF? Or is being the ED of the WMF your ideal job? Perhaps someday you’ll want to retire or just do something less stressful. What do you think you’ll do then?

KM: Luckily I really, truly love my job. Even on the tiredest mornings, I’m able to get out of bed feeling fortunate to do something I love, with people I respect and admire, on behalf of a mission I truly believe in. It can be a lot at times, but that’s the nature of working for a movement that has such a spectacularly aspirational vision for the world.

In the past, I always knew it was time to move on when I started being more interested in something outside of my work than inside the role. What I’ve found is that the beauty of Wikimedia is that everything is in the work, whether it’s meeting new people and partners, learning about places, languages, or cultures that are new to me, understanding new technologies and systems that are going to impact our lives, or ways that knowledge is used as a tool of power or empowerment. There is never a typical day, and I love that about us.

This is an incredible organization where I have the opportunity to take on an ambitious mission, and do so in partnership with a community of people around the world. It reflects a global generosity of spirit in a time when we need more of it. And maybe, if and when I do take a break, I’ll go sit on a beach somewhere for a bit… and find a promising article in need of a good edit!