Knowing False Negatives: An Adversarial Training Method for Distantly Supervised Relation Extraction
This addresses a specific issue in relation extraction for NLP researchers, offering an incremental improvement by handling false negatives more effectively.
The paper tackles the false negative problem in distantly supervised relation extraction, where incomplete knowledge bases cause mislabeled instances, and proposes a two-stage adversarial training method that identifies and utilizes these samples, achieving improved performance on two benchmark datasets.
Distantly supervised relation extraction (RE) automatically aligns unstructured text with relation instances in a knowledge base (KB). Due to the incompleteness of current KBs, sentences implying certain relations may be annotated as N/A instances, which causes the so-called false negative (FN) problem. Current RE methods usually overlook this problem, inducing improper biases in both training and testing procedures. To address this issue, we propose a two-stage approach. First, it finds out possible FN samples by heuristically leveraging the memory mechanism of deep neural networks. Then, it aligns those unlabeled data with the training data into a unified feature space by adversarial training to assign pseudo labels and further utilize the information contained in them. Experiments on two wildly-used benchmark datasets demonstrate the effectiveness of our approach.