Denoising Enhanced Distantly Supervised Ultrafine Entity Typing

TL;DR

This paper introduces a denoising approach for distantly supervised ultra-fine entity typing, using a noise model to improve label quality and an iterative training process to enhance entity typing accuracy.

Contribution

It proposes a novel noise estimation and denoising framework combined with a bi-encoder entity typing model trained iteratively for better performance.

Findings

Significantly outperforms baseline methods on Ultra-Fine dataset

Effective noise estimation improves label quality

Iterative training enhances entity typing accuracy

Abstract

Recently, the task of distantly supervised (DS) ultra-fine entity typing has received significant attention. However, DS data is noisy and often suffers from missing or wrong labeling issues resulting in low precision and low recall. This paper proposes a novel ultra-fine entity typing model with denoising capability. Specifically, we build a noise model to estimate the unknown labeling noise distribution over input contexts and noisy type labels. With the noise model, more trustworthy labels can be recovered by subtracting the estimated noise from the input. Furthermore, we propose an entity typing model, which adopts a bi-encoder architecture, is trained on the denoised data. Finally, the noise model and entity typing model are trained iteratively to enhance each other. We conduct extensive experiments on the Ultra-Fine entity typing dataset as well as OntoNotes dataset and demonstrate that our approach significantly outperforms other baseline methods.