CIL: Contrastive Instance Learning Framework for Distantly Supervised Relation Extraction

TL;DR

This paper introduces CIL, a contrastive instance learning framework that enhances distantly supervised relation extraction by better capturing sentence features and reducing noise, leading to improved performance over existing methods.

Contribution

The paper proposes a novel contrastive instance learning framework that extends beyond traditional multi-instance learning for relation extraction, effectively reducing noise and capturing richer sentence features.

Findings

Significant performance improvements on NYT10, GDS, and KBP datasets.

Effective reduction of noise in distantly supervised relation extraction.

Enhanced sentence feature representation through contrastive learning.

Abstract

The journey of reducing noise from distant supervision (DS) generated training data has been started since the DS was first introduced into the relation extraction (RE) task. For the past decade, researchers apply the multi-instance learning (MIL) framework to find the most reliable feature from a bag of sentences. Although the pattern of MIL bags can greatly reduce DS noise, it fails to represent many other useful sentence features in the datasets. In many cases, these sentence features can only be acquired by extra sentence-level human annotation with heavy costs. Therefore, the performance of distantly supervised RE models is bounded. In this paper, we go beyond typical MIL framework and propose a novel contrastive instance learning (CIL) framework. Specifically, we regard the initial MIL as the relational triple encoder and constraint positive pairs against negative pairs for each instance. Experiments demonstrate the effectiveness of our proposed framework, with significant improvements over the previous methods on NYT10, GDS and KBP.