A Review on Semi-Supervised Relation Extraction

TL;DR

This paper reviews semi-supervised relation extraction methods using deep learning and meta-learning, comparing self-ensembling, self-training, and dual learning approaches, highlighting their strengths and weaknesses.

Contribution

It provides a comprehensive comparison of three key semi-supervised RE methods and discusses representative models like Mean-teacher, LST, and DualRE.

Findings

Self-ensembling faces challenges with insufficient supervision.

Self-training iteratively improves by pseudo-labeling.

Dual learning leverages mutual feedback between tasks.

Abstract

Relation extraction (RE) plays an important role in extracting knowledge from unstructured text but requires a large amount of labeled corpus. To reduce the expensive annotation efforts, semisupervised learning aims to leverage both labeled and unlabeled data. In this paper, we review and compare three typical methods in semi-supervised RE with deep learning or meta-learning: self-ensembling, which forces consistent under perturbations but may confront insufficient supervision; self-training, which iteratively generates pseudo labels and retrain itself with the enlarged labeled set; dual learning, which leverages a primal task and a dual task to give mutual feedback. Mean-teacher (Tarvainen and Valpola, 2017), LST (Li et al., 2019), and DualRE (Lin et al., 2019) are elaborated as the representatives to alleviate the weakness of these three methods, respectively.