arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2205.09536cs.CL
收藏

一种简单有效的关系信息引导的Few-Shot关系抽取方法

A Simple yet Effective Relation Information Guided Approach for Few-Shot Relation Extraction

The Chinese University of Hong Kong (Shenzhen)(香港中文大学(深圳));Shenzhen Research Institute of Big Data(深圳大数据研究院);Pazhou Lab(琶洲实验室)

Yang Liu, Jinpeng Hu, Xiang Wan, Tsung-Hui Chang

摘要:

Few-Shot关系抽取的目标是通过在每个关系中使用一些标记的示例进行训练来预测句子中的一对实体的关系。一些最近的工作已经引入了关系信息(即,关系标签或描述)来辅助基于原型网络的模型学习。然而,大多数方法都是通过设计复杂的网络结构,如生成混合特征,结合对比学习或注意力网络等,用关系信息隐含地约束每个关系类的原型。我们认为,关系信息可以更明确和有效地引入到模型中。因此,本文提出了一种直接添加的方法来引入关系信息。具体地,对于每个关系类,首先通过连接关系的两个视图(即,[CLS]标记嵌入和所有标记的嵌入的平均值),然后直接添加到原始原型中用于训练和预测。在基准数据集FewRel 1.0上的实验结果显示出显着的改进,并取得了与最先进的结果相当的结果,这证明了我们提出的方法的有效性。此外,进一步的分析表明,直接添加是一个更有效的方式来集成的关系表示和原始原型。

英文摘要:

Few-Shot Relation Extraction aims at predicting the relation for a pair of entities in a sentence by training with a few labelled examples in each relation. Some recent works have introduced relation information (i.e., relation labels or descriptions) to assist model learning based on Prototype Network. However, most of them constrain the prototypes of each relation class implicitly with relation information, generally through designing complex network structures, like generating hybrid features, combining with contrastive learning or attention networks. We argue that relation information can be introduced more explicitly and effectively into the model. Thus, this paper proposes a direct addition approach to introduce relation information. Specifically, for each relation class, the relation representation is first generated by concatenating two views of relations (i.e., [CLS] token embedding and the mean value of embeddings of all tokens) and then directly added to the original prototype for both train and prediction. Experimental results on the benchmark dataset FewRel 1.0 show significant improvements and achieve comparable results to the state-of-the-art, which demonstrates the effectiveness of our proposed approach. Besides, further analyses verify that the direct addition is a much more effective way to integrate the relation representations and the original prototypes.

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。

↑