I Beg to Differ: A study of constructive disagreement in online conversations
This work addresses the challenge of understanding constructive disagreement in online communication, offering incremental improvements in prediction methods for content moderation.
The study tackled the problem of predicting whether online disagreements escalate to moderation by analyzing Wikipedia Talk page conversations, finding that neural models using conversation structure improved predictive accuracy over feature-based approaches.
Disagreements are pervasive in human communication. In this paper we investigate what makes disagreement constructive. To this end, we construct WikiDisputes, a corpus of 7 425 Wikipedia Talk page conversations that contain content disputes, and define the task of predicting whether disagreements will be escalated to mediation by a moderator. We evaluate feature-based models with linguistic markers from previous work, and demonstrate that their performance is improved by using features that capture changes in linguistic markers throughout the conversations, as opposed to averaged values. We develop a variety of neural models and show that taking into account the structure of the conversation improves predictive accuracy, exceeding that of feature-based models. We assess our best neural model in terms of both predictive accuracy and uncertainty by evaluating its behaviour when it is only exposed to the beginning of the conversation, finding that model accuracy improves and uncertainty reduces as models are exposed to more information.