DSGAN: Generative Adversarial Training for Distant Supervision Relation Extraction

Abstract

Distant supervision can effectively label data for relation extraction, but suffers from the noise labeling problem. Recent works mainly perform soft bag-level noise reduction strategies to find the relatively better samples in a sentence bag, which is suboptimal compared with making a hard decision of false positive samples in sentence level. In this paper, we introduce an adversarial learning framework, which we named DSGAN, to learn a sentence-level true-positive generator. Inspired by Generative Adversarial Networks, we regard the positive samples generated by the generator as the negative samples to train the discriminator. The optimal generator is obtained until the discrimination ability of the discriminator has the greatest decline. We adopt the generator to filter distant supervision training dataset and redistribute the false positive instances into the negative set, in which way to provide a cleaned dataset for relation classification. The experimental results show that the proposed strategy significantly improves the performance of distant supervision relation extraction comparing to state-of-the-art systems.

References in corpus (4)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Classifying Relations by Ranking with Convolutional Neural Networks
• Zero-Shot Learning -- The Good, the Bad and the Ugly
• KBGAN: Adversarial Learning for Knowledge Graph Embeddings

Cited by in corpus (4)

• Deep Neural Approaches to Relation Triplets Extraction: A Comprehensive Survey
• Neural Relation Extraction via Inner-Sentence Noise Reduction and Transfer Learning
• Relation Adversarial Network for Low Resource Knowledge Graph Completion
• Mention-centered Graph Neural Network for Document-level Relation Extraction