Zero-Shot Relation Triplet Extraction As Next-Sentence Prediction

Wenxiong Liao,Zhengliang Liu,Yiyang Zhang,Xiaoke Huang,Ninghao Liu,Tianming Liu,Quanzheng Li,Xiang Li,Hongmin Cai

DOI: https://doi.org/10.1016/j.knosys.2024.112507

2024-01-01

Abstract:Zero-shot relation triplet extraction (ZeroRTE) endeavors to extract relation triplets from a test set using a model trained on a training set with disjoint relations from the test set. Current ZeroRTE approaches primarily rely on two strategies: 1) Combining pre-trained language models to generate additional training samples; 2) Adding a large number of parameters that require training from scratch on top of a pre-trained language model. However, the former approach does not ensure the quality of generated samples, and the latter often struggles to generalize to unseen relations in the test set, particularly when the training set is small. In this paper, we introduce a novel method, Next Sentence Prediction for Relation Triplet Extraction (NSP-RTE), abstracting ZeroRTE as a higher-level next sentence prediction (NSP) task to enhance its generalization ability to unseen relation categories. NSP-RTE integrates modules for relation recognition, entity detection, and triplet classification, leveraging pre-trained BERT models with fewer parameters requiring training from scratch, while eliminating the need for additional sample generation. Our experiments on the FewRel and Wiki-ZSL datasets demonstrate that NSP-RTE, with its simple and efficient design, significantly outperforms previous methods.