SENT: Sentence-level Distant Relation Extraction via Negative Training

TL;DR

This paper introduces SENT, a novel sentence-level relation extraction framework that employs negative training to reduce noise from distant supervision, leading to improved accuracy and cleaner training data.

Contribution

The paper proposes a negative training approach integrated into SENT for more accurate sentence-level relation extraction from noisy distant supervision data.

Findings

Significant performance improvement over previous methods.

Effective noise filtering and data re-labeling.

Enhanced model robustness through negative training.

Abstract

Distant supervision for relation extraction provides uniform bag labels for each sentence inside the bag, while accurate sentence labels are important for downstream applications that need the exact relation type. Directly using bag labels for sentence-level training will introduce much noise, thus severely degrading performance. In this work, we propose the use of negative training (NT), in which a model is trained using complementary labels regarding that ``the instance does not belong to these complementary labels". Since the probability of selecting a true label as a complementary label is low, NT provides less noisy information. Furthermore, the model trained with NT is able to separate the noisy data from the training data. Based on NT, we propose a sentence-level framework, SENT, for distant relation extraction. SENT not only filters the noisy data to construct a cleaner dataset, but also performs a re-labeling process to transform the noisy data into useful training data, thus further benefiting the model's performance. Experimental results show the significant improvement of the proposed method over previous methods on sentence-level evaluation and de-noise effect.