Encouraging Emotion Regulation in Social Media Conversations through Self-Reflection

TL;DR

This paper presents a novel graph-based framework that promotes self-reflection in social media conversations, helping users regulate emotions and reduce toxicity by up to 12%, thus improving online well-being.

Contribution

It introduces a new technique for detecting emotional regulation needs and delivering didactic support through interaction design, emphasizing self-reflection and implicit regulation in online chats.

Findings

Can identify influential conversation threads related to toxicity

Helps reduce toxicity levels by up to 12%

First to focus on learning transfer via self-reflection in DER

Abstract

Anonymity in social media platforms keeps users hidden behind a keyboard. This absolves users of responsibility, allowing them to engage in online rage, hate speech, and other text-based toxicity that harms online well-being. Recent research in the field of Digital Emotion Regulation (DER) has revealed that indulgence in online toxicity can be a result of ineffective emotional regulation (ER). This, we believe, can be reduced by educating users about the consequences of their actions. Prior DER research has primarily focused on exploring digital emotion regulation practises, identifying emotion regulation using multimodal sensors, and encouraging users to act responsibly in online conversations. While these studies provide valuable insights into how users consciously utilise digital media for emotion regulation, they do not capture the contextual dynamics of emotion regulation online. Through interaction design, this work provides an intervention for the delivery of ER support. It introduces a novel technique for identifying the need for emotional regulation in online conversations and delivering information to users in a way that integrates didactic learning into their daily life. By fostering self-reflection in periods of intensified emotional expression, we present a graph-based framework for on-the-spot emotion regulation support in online conversations. Our findings suggest that using this model in a conversation can help identify its influential threads/nodes to locate where toxicity is concentrated and help reduce it by up to 12\%. This is the first study in the field of DER that focuses on learning transfer by inducing self-reflection and implicit emotion regulation.