Big in Japan - Wikimedia Commons

Big in Japan - Wikimedia Commons

Big in Japan Combating Systemic Bias Through Mobile Editing Oliver Keyes Research Analyst Wikimedia Foundation [email protected] (the increasing l...

549KB Sizes 0 Downloads 1 Views

Recommend Documents

Personal Computer - Wikimedia Commons
The PS/2's Model M keyboard from which modern PC ... two-handed alternatives more akin to a game controller, such as the

Cables v1.3 - Wikimedia Commons
Some computers and monitors. Some laptop computers. Some laptop computers. Aluminum Unibody MacBook (late 2008 - mid 200

WikiReader Internet - Wikimedia Commons
13.05.2004 - Viele dieser Browser (wie Mozilla, Internet Explorer) werden heu- ... Camino (früher: Chimera, für Mac OS

Inhaltsverzeichnis - Wikimedia Commons
Staaten, Hauptstädte und Landessprachen zuzuordnen. Flagge. Land. Hauptstadt. Landessprachen. (Minderheiten siehe. Fuß

Cloud Computing - Wikimedia Commons
Mar 10, 2015 - works: A revolution in the earth system science? ...... Grow to 26 Billion Units By 2020”. ... the Inte

Quizspiel - Wikimedia Commons
Absurdistan. Allwissende Müllhalde. Partymeile. Pommesgabel. Wikimedia logo family of Wikimedia Foundation. Wikipedia lo

Arduino - Wikimedia Commons
Jul 28, 2017 - Massimo Banzi (http://twit.tv/show/triangulation/110/) interviewed on the TV show Triangulation on the. T

Scientific mtti - Wikimedia Commons
with each other to co-operate i.n producing I others after It. The. oupling ... I pletely across a broad and deep river

nagarjunasagar dam - Wikimedia Commons
G. V. Gopalarao. Abstract. This article deals with the design and construction aspects of Nagarjunasagar. Dam situated i

Big in Japan Combating Systemic Bias Through Mobile Editing

Oliver Keyes Research Analyst Wikimedia Foundation [email protected] (the increasing line length is accidental) (alright, so now it's both deliberate and getting very silly)

/WHOIS

Got editing in 2005, 968 articles/77,787 edits since. ● Traffic analyst at the WMF. ● On top of that, the geodata specialist. ● ...also The Guy Who Knows Where The Bodies Live. ● “Oliver! We have a horrifying dataset. Your problem!” ●

User:ДиБгд, ru.wiki, 2007 https://commons.wikimedia.org/wiki/File:Brachiosaurus_DB.jpg

Ways of looking at Mobile

Getting more contributions from people who already edit. ● Dealing with systemic bias. ●

More contributions from existing people

If we're getting edits on desktop devices, we're only getting some editor time. ● At work ● ..pretending to work. ● At home and unwinding in front of the PC ●

More contributions from existing people By reducing the barrier to mobile contributions, we enable edits: ● On the train ● In front of the TV ● In bed ● More opportunities to edit = more edits! ●

Edits by hour , all usergroups, all projects >= 442155 observations per type

6%

% of requests

5%

type

4%

desktop mobile

3%

2%

1%

0

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Hour

Take-aways

I'll have a curry please! ● It's working! We get mobile contributions where we don't get desktop attention. ● Most people get really bored of their jobs starting around 1pm. ●

Combating systemic bias

A blocker to contributing to Wikipedia is access. ● By enabling more ways of accessing Wikipedia, we enable more ways of contributing to it. ● Mobile is such a way. ● How bad is the status quo? ●

“What is required... is a citizen of the world...reporting things of the day as if two thousand years in the past and those of the place they live in as if two thousand leagues distant.”

Diderot, Encyclopédie (1751)

Desktop− only editors, geolocated count 2000000

50

y

1500000 0

1000000 500000

− 50 − 100

0

x

100

Desktop edits, geolocated count

50

2000000

y

1500000 0

1000000 500000

− 50 − 100

0

x

100

Editors with exclusively mobile contributions, geolocated count

50

60000

y

40000 0 20000 − 50 − 100

0

x

100

Mobile edits, geolocated count 100000

50

y

75000 50000

0

25000 − 50 − 100

0

x

100

What about combating bias through content?

The distribution is a problem; we need people from different cultural backgrounds ● But maybe mobile editors edit in different areas and we can make up the difference there? ●

project Women's_History

U.S._counties

United_States

Texas

Russia

Religion

Politics

physics

Ohio

military_history

India

Food_and_drink

Comics

Cities

China

Chicago

Catholicism

Canada

California

Biography

biography

Australia

Albums

Album

Alabama

Freq 1200

prominent themes of mobile and desktop edits

900

type

600 desktop

mobile

300

0

Conclusion

We're doing a good job getting people to edit at different times thanks to mobile. ● We're not doing a good job getting people to edit different things. ● We're definitely not doing a good job getting different people to edit. ●

Mobile edits, geolocated count 100000

50

y

75000 50000

0

25000 − 50 − 100

0

x

100

How to help Go to mobile-l and tell the Mobile developers about the problems you face trying to contribute on Mobile; ● Write code! Patches, new features, better language support, kitchen.sink() ● Come up with research questions – how can we do this better? What are we missing? What should we be measuring? ● ...email them to me. I'm incredibly tired. ●

Credits Codebase, visualisations, datasets:

https://github.com/Ironholds/BigInJapan Polygon venting: Leila Zia Statistics venting: Adam Hyland Metadata venting: Dario Taraborelli Misc. venting: Kim Bruning, Thomas Fish, James Forrester, David Gerard, Aaron Halfaker, Katie Horn, Daniel Jones, Erica Litrenta, Tom Morris, Rohit Patnaik, that unrepentant reprobate, Emily Temple-Wood, Thehelpfulone, Courtney Thurston, Kat Walsh, Margaret Wander, Molly White. ...I do a lot of venting.

Questions?