HiCLRE: A Hierarchical Contrastive Learning Framework for Distantly Supervised Relation Extraction

TL;DR

HiCLRE introduces a hierarchical contrastive learning framework that effectively reduces noise in distantly supervised relation extraction by integrating multi-level interactions and dynamic data augmentation, leading to improved performance.

Contribution

It presents a novel three-level hierarchical contrastive learning framework with multi-granularity recontextualization and dynamic gradient-based augmentation for DSRE.

Findings

Significantly outperforms baseline methods on DSRE datasets.

Effective noise reduction through hierarchical interaction.

Improved relation extraction accuracy.

Abstract

Distant supervision assumes that any sentence containing the same entity pairs reflects identical relationships. Previous works of distantly supervised relation extraction (DSRE) task generally focus on sentence-level or bag-level de-noising techniques independently, neglecting the explicit interaction with cross levels. In this paper, we propose a hierarchical contrastive learning Framework for Distantly Supervised relation extraction (HiCLRE) to reduce noisy sentences, which integrate the global structural information and local fine-grained interaction. Specifically, we propose a three-level hierarchical learning framework to interact with cross levels, generating the de-noising context-aware representations via adapting the existing multi-head self-attention, named Multi-Granularity Recontextualization. Meanwhile, pseudo positive samples are also provided in the specific level for contrastive learning via a dynamic gradient-based data augmentation strategy, named Dynamic Gradient Adversarial Perturbation. Experiments demonstrate that HiCLRE significantly outperforms strong baselines in various mainstream DSRE datasets.