Learning Dual Retrieval Module for Semi-supervised Relation Extraction

By Hongtao Lin, Jun Yan, Meng Qu, Xiang Ren. 2019

67 citations recorded by the source corpus.

Relation extraction is an important task in structuring content of text data, and becomes especially challenging when learning with weak supervision-where only a limited number of labeled sentences are given and a large number of unlabeled sentences are available. Most existing work exploits unlabeled data based on the ideas of self-training (i.e., bootstrapping a model) and self-ensembling (e.g., ensembling multiple model variants). However, these methods either suffer from the issue of semantic drift, or do not fully capture the problem characteristics of relation extraction. In this paper, we leverage a key insight that retrieving sentences expressing a relation is a dual task of predicting the relation label for a given sentence-two tasks are complementary to each other and can be optimized jointly for mutual enhancement. To model this intuition, we propose DualRE, a principled framework that introduces a retrieval module which is jointly trained with the original relation prediction module. In this way, high-quality samples selected by the retrieval module from unlabeled data can be used to improve the prediction module, and vice versa. Experimental results1 on two public datasets as well as case studies demonstrate the effectiveness of the DualRE approach.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Zhiyuan Liu
2. Xiang Ren
3. Maosong Sun
4. Philip S. Yu
5. Xu Han
6. Bill Yuchen Lin
7. Tianyu Gao
8. Qinyuan Ye

Citations in the corpus

• Curriculum-Meta Learning for Order-Robust Continual Relation Extraction (2021, methodology): “2016; Miwa and Bansal 2016), semi-supervised methods (Chen et al. 2006; Sun, Grishman, and Sekine 2011; Lin et al. 2019), and distantly supervised methods (Yao et al.”
• Neural Snowball for Few-Shot Relation Learning (2019, background): “Many other semi-supervised methods can also be adopted for RE (Blum and Mitchell, 1998; Rosenberg et al., 2005; French et al., 2017; Lin et al., 2019), yet they still require sufficient labeled data and mainly aim at classifying predefined relations rather than discovering new relations.”
• Semi-supervised Relation Extraction via Incremental Meta Self-Training (2020, methodology): “For the evaluation metrics, following the previous work on relation extraction (Rosenberg, Hebert, and Schneiderman 2005; Tarvainen and Valpola 2017; Lin et al. 2019), we consider F1 score as the main metric while precision and recall serve as auxiliary metrics.”
• A Survey of Fake News (2020, unknown): “More recent studies in relation extraction can be seen in [Di et al. 2019; Lin et al. 2019; Yu et al. 2019].”
• NERO: A Neural Rule Grounding Framework for Label-Efficient Relation Extraction (2020, background): “Specifically, here we focus on sentence-level RE [11, 18, 38], which aims to predict the relation of entity mention pairs in a sentence s—i.”
• Learning from Explanations with Neural Execution Tree (2019, unknown): “(4) DualRE (Lin et al., 2019) jointly trains a relation prediction module and a retrieval module.”
• Learning to Annotate: Modularizing Data Augmentation for Text Classifiers with Natural Language Explanations (2019, methodology): “(4) DualRE (Lin et al., 2019) jointly trains a relation prediction module and a retrieval module.”