Re-entry Prediction for Online Conversations via Self-Supervised Learning

TL;DR

This paper introduces a self-supervised learning approach for re-entry prediction in online conversations, leveraging conversation patterns and user engagement signals to improve prediction accuracy.

Contribution

It proposes three novel auxiliary tasks for self-supervised learning that enhance re-entry prediction beyond traditional history-based methods.

Findings

Outperforms previous state-of-the-art methods

Requires fewer parameters and converges faster

Effective across datasets from Twitter and Reddit

Abstract

In recent years, world business in online discussions and opinion sharing on social media is booming. Re-entry prediction task is thus proposed to help people keep track of the discussions which they wish to continue. Nevertheless, existing works only focus on exploiting chatting history and context information, and ignore the potential useful learning signals underlying conversation data, such as conversation thread patterns and repeated engagement of target users, which help better understand the behavior of target users in conversations. In this paper, we propose three interesting and well-founded auxiliary tasks, namely, Spread Pattern, Repeated Target user, and Turn Authorship, as the self-supervised signals for re-entry prediction. These auxiliary tasks are trained together with the main task in a multi-task manner. Experimental results on two datasets newly collected from Twitter and Reddit show that our method outperforms the previous state-of-the-arts with fewer parameters and faster convergence. Extensive experiments and analysis show the effectiveness of our proposed models and also point out some key ideas in designing self-supervised tasks.