H-FND: Hierarchical False-Negative Denoising for Distant Supervision Relation Extraction

12/07/2020
by   Jhih-Wei Chen, et al.
0

Although distant supervision automatically generates training data for relation extraction, it also introduces false-positive (FP) and false-negative (FN) training instances to the generated datasets. Whereas both types of errors degrade the final model performance, previous work on distant supervision denoising focuses more on suppressing FP noise and less on resolving the FN problem. We here propose H-FND, a hierarchical false-negative denoising framework for robust distant supervision relation extraction, as an FN denoising solution. H-FND uses a hierarchical policy which first determines whether non-relation (NA) instances should be kept, discarded, or revised during the training process. For those learning instances which are to be revised, the policy further reassigns them appropriate relations, making them better training inputs. Experiments on SemEval-2010 and TACRED were conducted with controlled FN ratios that randomly turn the relations of training and validation instances into negatives to generate FN instances. In this setting, H-FND can revise FN instances correctly and maintains high F1 scores even when 50 further conducted to shows that H-FND is applicable in a realistic setting.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/30/2020

RDSGAN: Rank-based Distant Supervision Relation Extraction with Generative Adversarial Framework

Distant supervision has been widely used for relation extraction but suf...
research
06/08/2023

Open Set Relation Extraction via Unknown-Aware Training

The existing supervised relation extraction methods have achieved impres...
research
05/21/2021

Revisiting the Negative Data of Distantly Supervised Relation Extraction

Distantly supervision automatically generates plenty of training samples...
research
08/12/2020

Distantly Supervised Relation Extraction in Federated Settings

This paper investigates distantly supervised relation extraction in fede...
research
04/17/2022

Does Recommend-Revise Produce Reliable Annotations? An Analysis on Missing Instances in DocRED

DocRED is a widely used dataset for document-level relation extraction. ...
research
11/14/2017

False Positive and Cross-relation Signals in Distant Supervision Data

Distant supervision (DS) is a well-established method for relation extra...
research
05/24/2018

Robust Distant Supervision Relation Extraction via Deep Reinforcement Learning

Distant supervision has become the standard method for relation extracti...

Please sign up or login with your details

Forgot password? Click here to reset