Improving Distantly Supervised Relation Extraction with Self-Ensemble Noise Filtering

TL;DR

This paper introduces a self-ensemble filtering method to reduce noise in distantly supervised relation extraction, improving model robustness and F1 scores on the NYT dataset.

Contribution

It proposes a novel self-ensemble filtering mechanism to effectively remove noisy samples during training in distantly supervised relation extraction.

Findings

Improves F1 scores of neural relation extraction models.

Enhances robustness of models against noisy training data.

Validated on NYT dataset with multiple state-of-the-art models.

Abstract

Distantly supervised models are very popular for relation extraction since we can obtain a large amount of training data using the distant supervision method without human annotation. In distant supervision, a sentence is considered as a source of a tuple if the sentence contains both entities of the tuple. However, this condition is too permissive and does not guarantee the presence of relevant relation-specific information in the sentence. As such, distantly supervised training data contains much noise which adversely affects the performance of the models. In this paper, we propose a self-ensemble filtering mechanism to filter out the noisy samples during the training process. We evaluate our proposed framework on the New York Times dataset which is obtained via distant supervision. Our experiments with multiple state-of-the-art neural relation extraction models show that our proposed filtering mechanism improves the robustness of the models and increases their F1 scores.