TemPrompt: Multi-Task Prompt Learning for Temporal Relation Extraction in RAG-based Crowdsourcing Systems
This work addresses data scarcity issues in temporal relation extraction for crowdsourcing systems, offering an incremental improvement through novel prompt construction and auxiliary tasks.
The paper tackles the problem of limited and unevenly distributed annotated data in temporal relation extraction (TRE) for crowdsourcing systems by proposing TemPrompt, a multi-task prompt learning framework that incorporates prompt tuning and contrastive learning. The results show that TemPrompt outperforms baselines across most metrics in standard and few-shot settings, with a case study validating its effectiveness in printed circuit board design and manufacturing scenarios.
Temporal relation extraction (TRE) aims to grasp the evolution of events or actions, and thus shape the workflow of associated tasks, so it holds promise in helping understand task requests initiated by requesters in crowdsourcing systems. However, existing methods still struggle with limited and unevenly distributed annotated data. Therefore, inspired by the abundant global knowledge stored within pre-trained language models (PLMs), we propose a multi-task prompt learning framework for TRE (TemPrompt), incorporating prompt tuning and contrastive learning to tackle these issues. To elicit more effective prompts for PLMs, we introduce a task-oriented prompt construction approach that thoroughly takes the myriad factors of TRE into consideration for automatic prompt generation. In addition, we design temporal event reasoning in the form of masked language modeling as auxiliary tasks to bolster the model's focus on events and temporal cues. The experimental results demonstrate that TemPrompt outperforms all compared baselines across the majority of metrics under both standard and few-shot settings. A case study on designing and manufacturing printed circuit boards is provided to validate its effectiveness in crowdsourcing scenarios.