, 17 tweets, 9 min read Read on Twitter
While I'm at it, here are some of my favorite sources of empirical information in this area.
1. All the empirical studies I know about "over-removal" in legal takedowns are listed here: cyberlaw.stanford.edu/blog/2015/10/e…
2. The best of them is Urban et al 2016. I summarize it here cyberlaw.stanford.edu/blog/2016/04/d…
4. Facebook put out a big, multi-part info dump on its content moderation practices in 2018. It's not perfect but it was a big step forward
transparency.facebook.com/community-stan…
fbnewsroomus.files.wordpress.com/2018/05/unders…
facebook.com/communitystand…
5. YouTube put out a more robust transparency report in 2018, too transparencyreport.google.com/youtube-policy…
6. Google's data dump about three years of Right to Be Forgotten removals is super useful, both for the raw data it shows (highly aggregated, for legal reasons) and as a model/strawman for thinking about what level of granularity is reasonable to expect. transparencyreport.google.com/eu-privacy/ove….
7. The in-depth human rights audit of the Internet Watch Foundation in the UK is factual and super illuminating about the complexities of assessing content at scale, balancing state and private power, and protecting human rights. iwf.org.uk/sites/default/…
8. Harvard's @lumendatabase is an unparalleled resource for raw information about takedown demands: lumendatabase.org
9. A deep dive on another model - Reddit's diverse, forum-specific, norm- and user-drive content moderation: eegilbert.org/papers/cscw18-…
10. What companies do transparency reports, who is doing a good job, and how could they do even better? Check out:
eff.org/who-has-your-b…
newamerica.org/oti/blog/annou…
accessnow.org/transparency-r….
11. Are content filters good at identifying forbidden expression (without smacking down lawful/permitted speech at the same time)? There is woefully little research out there, even as filtering spreads. The best I know are:
engine.is/the-limits-of-…
cdt.org/insight/mixed-…
12. Do content takedowns disproportionately impact vulnerable and minority groups, are policies enforced inconsistently on the basis of race, etc.? Anecdotes abound, but empirical research is terribly hard for lack of transparency. onlinecensorship.org has reports, though.
13. There is simply no replacement for talking to the people who do content moderation in real life. @ericgoldman's groundbreaking COMO conference did just that. law.scu.edu/event/content-…
Similar events in DC and NYC built on it, as will @AlexandraQu's in Brussels & another in DC.
14. What do real people do when they think their speech and clicks are being monitored for wrongdoing? The curtail speech and research on legal and important topics. papers.ssrn.com/sol3/papers.cf…
papers.ssrn.com/sol3/papers.cf…
pen.org/chilling-effec…
15. What skullduggery can you find by digging around in the @lumendatabase? This story's still unfolding, but @VolokhC found people falsifying court orders, or misleading courts into issuing real ones, in order to get Google to take things down. washingtonpost.com/news/volokh-co…
11. More is needed. Here is my list of key questions for transparency and new research. We can't make sound policy when these things are left to anecdata. There is so much great work to be done!
Just realized I got up to item 15, and then numbered the next one 11. Sorry folks. I'm usually better with numbers than that.
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Daphne Keller
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Follow Us on Twitter!

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!