Event Guided Denoising for Multilingual Relation Learning

TL;DR

This paper introduces a data collection method for relation extraction that leverages news article structures to create a high-quality, multilingual training corpus, achieving state-of-the-art results with significantly less data and computational cost.

Contribution

The authors propose a novel denoising approach for multilingual relation learning that reduces training data requirements while maintaining competitive performance.

Findings

Comparable performance to state-of-the-art with fewer examples

Effective multilingual relation extraction in English and Spanish

Reduced training cost and data requirements

Abstract

General purpose relation extraction has recently seen considerable gains in part due to a massively data-intensive distant supervision technique from Soares et al. (2019) that produces state-of-the-art results across many benchmarks. In this work, we present a methodology for collecting high quality training data for relation extraction from unlabeled text that achieves a near-recreation of their zero-shot and few-shot results at a fraction of the training cost. Our approach exploits the predictable distributional structure of date-marked news articles to build a denoised corpus -- the extraction process filters out low quality examples. We show that a smaller multilingual encoder trained on this corpus performs comparably to the current state-of-the-art (when both receive little to no fine-tuning) on few-shot and standard relation benchmarks in English and Spanish despite using many fewer examples (50k vs. 300mil+).