Continuous Detection, Rapidly React: Unseen Rumors Detection based on Continual Prompt-Tuning

TL;DR

This paper introduces CPT-RD, a continual prompt-tuning framework for rumor detection on social platforms that adapts to evolving data streams without forgetting past knowledge, enabling rapid response to emerging rumors.

Contribution

It proposes a novel continual learning approach with task-conditioned prompt hypernetworks that prevents catastrophic forgetting and facilitates knowledge transfer across domains in rumor detection.

Findings

Effective in handling sequential rumor detection tasks

Avoids catastrophic forgetting without rehearsal buffer

Enhances bidirectional knowledge transfer

Abstract

Since open social platforms allow for a large and continuous flow of unverified information, rumors can emerge unexpectedly and spread quickly. However, existing rumor detection (RD) models often assume the same training and testing distributions and can not cope with the continuously changing social network environment. This paper proposed a Continual Prompt-Tuning RD (CPT-RD) framework, which avoids catastrophic forgetting (CF) of upstream tasks during sequential task learning and enables bidirectional knowledge transfer between domain tasks. Specifically, we propose the following strategies: (a) Our design explicitly decouples shared and domain-specific knowledge, thus reducing the interference among different domains during optimization; (b) Several technologies aim to transfer knowledge of upstream tasks to deal with emergencies; (c) A task-conditioned prompt-wise hypernetwork (TPHNet) is used to consolidate past domains. In addition, CPT-RD avoids CF without the necessity of a rehearsal buffer.