In depth predictive statistical analysis of troll like anti-social behavior for potential future banned users (FBUs) ...
Characterizing antisocial behavior.
We compare the activity of users who are later banned from a community, or
Future-Banned Users (FBUs), with that of users who were
never banned, or Never-Banned Users (NBUs).
By analyzing the language of their posts, we find significant differences
between these two groups. For example, FBUs tend to write
less similarly to other users, and their posts are harder to un-
derstand according to standard readability metrics. They are
also more likely to use language that may stir further conflict
(e.g., they use less positive words and use more profanity).http://arxiv.org/pdf/1504.00680v1.pdfPredicting future banning.
Last, we show that a user’s posting behavior can be used to make predictions about who
will be banned in the future. Inspired by our empirical analysis,
we design features that capture various aspects of antisocial behavior: post content, user activity, community response,
and the actions of community moderators. We find that we can predict with over 80% AUC (area under the ROC
curve) whether a user will be subsequently banned.
In fact,
we only need to observe 5 to 10 user’s posts before a classifier is able to make a reliable prediction. Further, cross-
domain classification performance remains high, suggesting that the features indicative of antisocial behavior
that we discover are not community-specific.
