Detecting Evil Through Machine Learning

Reddit Computer Science, 2016-17

Liaison(s): (not listed)
Advisor(s): Yi-Chieh Wu
Students(s): Jonathan Chang (PM-S), Rachel Lee, Anna Ma, Kent Shikama (PM-F), Lisa Yin

Reddit is an online discussion platform where users can form communities centered around a variety of topics. Like any discussion website, Reddit has experienced instances of spam, trolling, cyberbullying, and general aggressive behaviors from users, which can alienate other users or dis-courage people from joining Reddit. To help Reddit combat such behavior, we created an extensible machine learning pipeline for predicting whether a comment will be perceived as aggressive.