Antisocial Computing

8480b47e733a040fba07c32da414b0e0?s=47 Justin Cheng
February 02, 2018

Antisocial Computing

8480b47e733a040fba07c32da414b0e0?s=128

Justin Cheng

February 02, 2018
Tweet

Transcript

  1. Justin Cheng / @jcccf / clr3.com with Jure Leskovec, Michael

    Bernstein, Cristian Danescu-Niculescu-Mizil, and Tim Hsieh ANTISOCIAL COMPUTING Explaining, Predicting, and Mediating Online Negative Behavior v1.1
  2. Vieweg, et al. (2010); Kittur, et al. (2013); Burke &

    Kraut (2016)
  3. None
  4. 47% of online users have been harassed Data & Society

    (2017)
  5. Popular Science (2013); The Verge (2015); Chicago Sun-Times (2014)

  6. Why is bad behavior so prevalent?
 (╯°□°)╯︵ ┻━┻ Research Question

  7. Understanding bad behavior helps us build healthier communities Implications Systems

    Guidelines Interventions
  8. Antisocial behavior is largely due to sociopaths Prior Work Donath

    (1999); Hardaker (2010); Buckels, et al. (2014)
  9. Antisocial behavior is largely due to ordinary people This Work

  10. Antisocial Computing Talk Outline What causes antisocial behavior? How do

    systems mediate it? Does it worsen over time? 1 2 3 WIP!
  11. Antisocial Computing Talk Outline What causes antisocial behavior? How do

    systems mediate it? Does it worsen over time? 1 2 3 WIP!
  12. CSCW 2017 (Best Paper); ICWSM 2015 (Honorable Mention) with M.

    Bernstein, C. Danescu-Niculescu-Mizil, J. Leskovec CAN ANYONE BECOME A TROLL? Causes of Antisocial Behavior in Online Discussions
  13. CONTENT WARNING!
 This talk contains depictions of trolling that use

    strong language. !
  14. None
  15. It also shows that Islam and Christianity teaching women to

    dress modest could be right afterall.
  16. It also shows that Islam and Christianity teaching women to

    dress modest could be right afterall. Clearly that is the only logical conclusion to this article. Now if you'll excuse me, I need to iron my tarp. I have work on Monday, and I want to appear 'modest'. fail at life. go bomb yourself. Religious nut alert
  17. We studied multiple large comment- based news communities. 470M posts

    831M votes 76M users
  18. What is trolling?

  19. What is trolling? Engaging in negatively marked online behavior Taking

    pleasure in upsetting others Not following the rules Disrupting a group while staying undercover Donath (1999); Hardaker (2010); Kirman (2012); Schwartz (2008)
  20. Trolling is behavior that occurs outside community norms. Defined using

    community guidelines Our Definition e.g., name-calling, personal attacks, profanity, threats, hate speech, ethnically/racially offensive material
  21. Are trolls just a vocal minority? Donath (1999); Hardaker (2010);

    Shachaf & Hara (2010); NYT (2008); Wired (2014); Vox (2014)
  22. How much do trolls troll? Proportion of Banned Users 0

    0.1 0.2 0.3 0.4 Proportion of Deleted Posts 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  23. The distribution of trolls is bimodal Proportion of Banned Users

    0 0.1 0.2 0.3 0.4 Proportion of Deleted Posts 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  24. Are there two types of trolls? Proportion of Banned Users

    0 0.1 0.2 0.3 0.4 Proportion of Deleted Posts 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Situational trolling? Lifelong trolling?
  25. What if antisocial behavior is situational?

  26. Challenge: how to show that antisocial behavior is situational? Observational

    data isn’t causal
  27. Challenge: how to show that antisocial behavior is situational? Experiments

    hard to generalize
  28. Simulated Discussion Experiment Large-Scale Analysis Solution: Experiment + Observational Study

  29. Anyone can become a troll Our Hypothesis

  30. “Broken windows” theory Zimbardo (1969); Wilson (1982)

  31. Unpleasant stimuli increase aggression Jones & Bogat (1978); Rotton &

    Frey (1985)
  32. Experiment: simulated discussion forum

  33. N=667, 40% female Quiz Discussion Experimental method

  34. Quiz Discussion Experimental method ×

  35. Positive/Negative Mood Positive/Negative Context Experimental method × Quiz Discussion

  36. Easy quiz (positive mood)

  37. Difficult quiz (negative mood)

  38. Positive discussion context

  39. Negative discussion context

  40. How did trolling differ across conditions? Two expert raters labeled

    posts independently
  41. How did trolling differ across conditions? Positive Mood Negative Mood

    Positive Context Negative Context % Troll Posts
  42. Trolling is lowest with positive conditions… Positive Mood Negative Mood

    Positive Context 35% Negative Context % Troll Posts
  43. …increases with either negative condition… Positive Mood Negative Mood Positive

    Context 35% 49% Negative Context 47% % Troll Posts
  44. …and almost doubles in the worst case Positive Mood Negative

    Mood Positive Context 35% 49% Negative Context 47% 68% % Troll Posts (p < 0.05 using a mixed effects logistic regression model)
  45. Negative affect almost triples Positive Mood Negative Mood Positive Context

    1.1% 1.4% Negative Context 2.3% 2.9% % Negative Affect Words (LIWC) (p < 0.05)
  46. Hilary is a solid candidate. As a woman, I appreciate

    that she's a woman, but it's not the only reason I think she would do well in office. Positive Mood + Context
  47. Anyone who votes for her is a complete idiot. These

    supporters are why this country is in such bad shape now. Uneducated people. Negative Mood + Context
  48. Bad mood and negative discussion context increase trolling

  49. Simulated Discussion Experiment Large-Scale Analysis of CNN.com Online Experiment +

    Observational Study
  50. Can trolling, like mood, vary with the time of day

    and day of week? Replicating Mood Golder & Macy (2011) Neg. Affect Time of Day ?
  51. How does trolling vary with time of day? Proportion of

    Flagged Posts 0.03 0.033 0.036 0.039 0.042 Time of Day 0 6 12 18 24
  52. Trolling peaks in the evening… Proportion of Flagged Posts 0.03

    0.033 0.036 0.039 0.042 Time of Day 0 6 12 18 24 Negative Affect (Golder & Macy)
  53. …and early in the work week. Proportion of Flagged Posts

    0.03 0.032 0.034 0.036 0.038 0.04 Day of Week Mon Tue Wed Thu Fri Sat Sun
  54. Trolling peaks when moods are worse Time of Day Proportion

    of Flagged Posts Negative Affect Proportion of Downvotes Day of Week
  55. Mood spills over from prior discussions Replicating Mood

  56. … … (Discussion) Mood spills over from prior discussions

  57. … … … ? … (Discussion) (Unrelated Discussions) … ?

    … Mood spills over from prior discussions
  58. … … … … (Discussion) … … Mood spills over

    from prior discussions
  59. A user who trolled in a previous discussion is twice

    as likely to troll in a later, unrelated discussion (p < 0.01) Replicating Mood
  60. The initial post affects subsequent trolling Replicating Context

  61. (Separate discussions of same article) … ? ? ? …

    ? ? ? The initial post affects subsequent trolling
  62. (Separate discussions of same article) … … The initial post

    affects subsequent trolling
  63. An initial troll post increases the subsequent trolling by 63%

    (p < 0.01) Replicating Context
  64. Troll or not? User

  65. Troll or not? User Mood

  66. Troll or not? User Mood Other users { }

  67. Because trolling is situational, ordinary people can end up trolling

  68. None
  69. Can voting mitigate bad behavior?

  70. Downvoting causes negative behavior to worsen Our Hypothesis

  71. Antisocial Computing Talk Outline What causes antisocial behavior? How do

    systems mediate it? Does it worsen over time? 1 2 3 WIP!
  72. How Antisocial Behavior Worsens ICWSM 2014 with C. Danescu-Niculescu-Mizil, J.

    Leskovec CAN ANTISOCIAL BEHAVIOR SPI R A L?
  73. Downvoting causes negative behavior to worsen Our Hypothesis

  74. What effects do evaluations have? Positively evaluated ? ? Negatively

    evaluated
  75. What is a positive or negative evaluation?

  76. Defining positive and negative evaluations : 9 1 2 8

    N↑
 N↑ + N↓ 2
 2+8 = = 0.2 p↓ ≤ : N↑
 N↑ + N↓ 9
 9+1 = = 0.9 p↑ ≥ Positive Evaluation Negative Evaluation (validated using a crowdsourcing experiment)
  77. What effects do evaluations have? Positively evaluated ? ? Negatively

    evaluated
  78. Does feedback encourage better behavior? Skinner (1938)

  79. Or is bad stronger than good? Brinko (1993); Baumeister, et

    al. (2001)
  80. Four large comment-based news communities

  81. What effects do evaluations have? Positively evaluated Negatively evaluated

  82. What effects do evaluations have? … … … …

  83. What effects do evaluations have? … … … … Before

    After vs. Before After vs.
  84. Challenge: how to compare different users and posts? Aren’t downvoted

    users/posts inherently worse?
  85. Solution: propensity score matching PSM: Rosenbaum (1983); CEM: Iacus, et

    al. (2012) Positively evaluated Negatively evaluated
  86. Match on text quality Similar text quality q } ≈

  87. Computing text quality (Validated using crowdsourcing) Learn p with bigrams

    1 3 lorem ipsum… q = ? Lorem… ? ? 9 2 lorem ipsum… … Text quality q is predicted p
  88. Match on text quality Similar text quality q(c↑ )=q(c↓ )

    } ≈
  89. …as well as other covariates Similar history (# posts, overall

    proportion of upvotes, etc.) { ≈ … … ≈
  90. …as well as other covariates ≈ … … … …

  91. How are subsequent posts evaluated? ≈ … … … …

  92. How much are evaluations due to textual or community effects?

  93. How much are evaluations due to textual effects (i.e., people

    writing worse)? f***ing a****** i.e., downvoting because of post content
  94. How much are evaluations due to community effects (i.e., inherent

    bias)? We dislike you. i.e., downvoting because of community dislikes author We dislike you.
  95. Do people write better/worse after a positive/negative evaluation? Textual Effects

  96. ≈ … … … … Better/Worse? Do people write better/worse

    after a positive/negative evaluation? ≈
  97. Text quality drops significantly after a negative evaluation… (p <

    0.05, mean effect size r = 0.18) … … Negativity bias
  98. …but doesn’t change after a positive evaluation … … (n.s.)

    Negativity bias
  99. How does community bias change after an evaluation? Community Effects

  100. … … … … How does community bias change after

    an evaluation?
  101. Measuring community bias N↑ q N↓ N↑
 N↑ + N↓

    = 0.5 p(c) = = 0.8 q(c) p(c) q(c) = 0.3 Prop. Upvotes Text Quality Community Bias − −
  102. Community bias increase more after a negative than positive evaluation

    (p < 0.01, mean effect size r = 0.13) … … Halo effect
  103. Troll or not? User Mood Other users { }

  104. Troll or not? User Mood Other users { }

  105. Troll or not? User Mood Other users { }

  106. Trolls may start out normal, but tip into a spiral

    and never recover
  107. Were downvotes a good idea in the first place?

  108. Antisocial Computing Talk Outline What causes antisocial behavior? How do

    systems mediate it? Does it worsen over time? 1 2 3 WIP!
  109. Work-in-Progress with J. Leskovec, T. Hsieh WIP! How Introducing Downvoting

    to Communities Impacts User Behavior CAN ENABLING NEGATIVE FEEDBACK HARM COMMUNITIES?
  110. Downvoting causes negative behavior to worsen… Previously WIP!

  111. …but does its mere presence affect communities? The Present Work

    WIP!
  112. The ability to downvote negatively alters behavior Our Hypothesis WIP!

  113. Challenge: how do we measure the impact of introducing downvoting?

    WIP!
  114. Why the sudden increase in downvotes? Data: Worldstarhiphop circa. 2012

    # Votes 0 40000 80000 120000 160000 Time (Days) 0 20 40 60 80 100 120 140 160 180 200 220 Downvotes Upvotes ? WIP!
  115. Early 2012 Late 2012 Cause: a Disqus interface change WIP!

  116. Early 2012 Late 2012 Cause: a Disqus interface change +

    WIP!
  117. A natural experiment studying the impact of introducing downvoting Opportunity

  118. Twenty comment-based websites … … WIP!

  119. How do we measure impact? WIP! Time (Weeks) 0

  120. How do we measure impact? WIP! Time (Weeks) 1 0

  121. How do we measure impact? WIP! Time (Weeks) 1 5

    -4 0 Before After Repeat for each domain (x20), applying corrections as necessary
  122. How did the introduction of downvoting affect user behavior? Votes

    Discussions Anonymity WIP!
  123. Does introducing downvoting alter upvoting behavior? WIP!

  124. Does introducing downvoting alter upvoting behavior? ? ? Downvotes don’t

    impact upvotes overall Upvotes increase to offset downvotes WIP! (Muchnik, et al. 2013)
  125. Introducing downvoting has little effect on upvoting overall. (5 domains

    saw a significant increase / 12 saw no change / 3 saw a significant decrease, overall n.s.) WIP!
  126. Does introducing downvoting spur discussion? WIP!

  127. Does introducing downvoting spur discussion? ? ? Downvotes cause arguments

    Downvotes replace comments WIP!
  128. Introducing downvoting spurs discussion. More replies, less top-level comments (13

    + / 4 · / 3 —, overall increase [p < 0.01]) WIP!
  129. Introducing downvoting spurs discussion. More replies, less top-level comments Discussion

    length increases … (13 + / 4 · / 3 —, overall increase [p < 0.01]) (15 + / 3 · / 2 —, overall increase [p < 0.01]) WIP!
  130. Introducing downvoting spurs discussion. More replies, less top-level comments Discussion

    length increases More back-and-forth discussion … @john… @mary… @john… @mary… (13 + / 4 · / 3 —, overall increase [p < 0.01]) (15 + / 3 · / 2 —, overall increase [p < 0.01]) (16 + / 2 · / 2 —, overall increase [p < 0.01]) WIP!
  131. Does introducing downvoting have a “chilling effect”? WIP!

  132. Introducing downvoting increases anonymity. WIP!

  133. Introducing downvoting increases anonymity. WIP! (18 sig. increase / 1

    no change / 1 sig. decrease, overall increase [p < 0.01])
  134. Negative feedback mechanisms can (unintentionally) encourage trolling

  135. What we now know What we thought Trolls are a

    vocal minority Trolls can be ordinary people Trolling is innate Trolling can spiral from a single bad post Design mediates trolling behavior Design mediates a user’s experience ANTISOCIAL COMPUTING
  136. Justin Cheng / @jcccf / clr3.com with Jure Leskovec, Michael

    Bernstein, Cristian Danescu-Niculescu-Mizil, and Tim Hsieh ANTISOCIAL COMPUTING Explaining, Predicting, and Mediating Online Negative Behavior v1.1