arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2302.08068cs.CLcs.AIcs.IRcs.LG
收藏

LabelPrompt:面向关系分类的高效提示学习

LabelPrompt: Effective Prompt-based Learning for Relation Classification

Jiangnan University(江南大学);University of Surrey(萨里大学)

Wenjie Zhang, Xiaoning Song, Zhenhua Feng, Tianyang Xu, Xiaojun Wu

AI总结:

本文提出LabelPrompt提示学习框架,通过关系标签词元、实体感知对比学习和注意力查询策略提升关系分类性能,在少样本基准实验中表现优越。

摘要:

近年来,基于提示的学习通过将自然语言处理(NLP)任务重构为完形填空式格式,使预训练语言模型(PLM)与下游任务更好地对齐,因而在众多NLP任务中受到欢迎。然而,将该方法应用于关系分类会面临独特挑战。具体而言,将填充掩码词元的自然语言词与语义关系标签(例如“org:founded_by”)关联起来十分困难。为应对这一挑战,本文提出一种名为LabelPrompt的新型基于提示的学习方法,用于关系分类任务。受“给模型选择!”这一直觉启发,我们首先定义额外词元来表示关系标签,将这些词元视为具有语义初始化的表述器,并使用提示模板方法显式构建它们。随后,为缓解预测关系与给定实体之间的不一致,我们通过对比学习实现了一个实体感知模块。最后,我们在自注意力层内实施注意力查询策略,以区分提示词元和序列词元。这些策略共同增强了基于提示学习的适应性,尤其是在仅有少量标注数据集可用时。在基准数据集上的综合实验证明了本方法的优越性,特别是在少样本场景中。

英文摘要:

Recently, prompt-based learning has gained popularity across many natural language processing (NLP) tasks by reformulating them into a cloze-style format to better align pre-trained language models (PLMs) with downstream tasks. However, applying this approach to relation classification poses unique challenges. Specifically, associating natural language words that fill the masked token with semantic relation labels (\textit{e.g.} \textit{``org:founded\_by}'') is difficult. To address this challenge, this paper presents a novel prompt-based learning method, namely LabelPrompt, for the relation classification task. Motivated by the intuition to ``GIVE MODEL CHOICES!'', we first define additional tokens to represent relation labels, which regard these tokens as the verbaliser with semantic initialisation and explicitly construct them with a prompt template method. Then, to mitigate inconsistency between predicted relations and given entities, we implement an entity-aware module with contrastive learning. Last, we conduct an attention query strategy within the self-attention layer to differentiates prompt tokens and sequence tokens. Together, these strategies enhance the adaptability of prompt-based learning, especially when only small labelled datasets is available. Comprehensive experiments on benchmark datasets demonstrate the superiority of our method, particularly in the few-shot scenario.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑