The good, the bad and the deviant in community question answering

Abstract Community question answering (CQA) are collaborative online places where members ask questions for others to answer. Community members on these platforms share their expertise on various topics, from mechanical repairs to parenting. As a crowd-sourced service, such platforms not only depend on user-provided questions and answers, but also rely on their users for monitoring and flagging content that violates community rules. This study focuses on user-reported flags to characterize the behavior of the good guys and bad guys in a popular community question answering, Yahoo Answers. Conventional wisdom is to eliminate the users who receive many flags. However, our analysis of a year of traces from Yahoo Answers shows that the number of flags does not tell the full story: on one hand, users with many flags may still contribute positively to the community. On the other hand, users who never get flagged are found to violate community rules and get their accounts suspended. This analysis, however, also shows that abusive users are betrayed by their network properties: we find strong evidence of homophilous behavior and use this finding to detect abusive users who go under the community radar. Based on our empirical observations, we build a classifier that is able to detect abusive users with an accuracy as high as 83%.