An End-to-end Model for Entity-level Relation Extraction using Multi-instance Learning

TL;DR

This paper introduces an end-to-end entity-level relation extraction model that leverages multi-instance learning and multi-level representations, achieving state-of-the-art results on the DocRED dataset.

Contribution

The paper presents a novel joint model for entity-level relation extraction that operates without mention-level annotations, combining coreference resolution with multi-instance learning.

Findings

Achieves state-of-the-art results on DocRED dataset.

First to report entity-level end-to-end relation extraction results.

Joint approach is as effective as task-specific methods, with improved efficiency.

Abstract

We present a joint model for entity-level relation extraction from documents. In contrast to other approaches - which focus on local intra-sentence mention pairs and thus require annotations on mention level - our model operates on entity level. To do so, a multi-task approach is followed that builds upon coreference resolution and gathers relevant signals via multi-instance learning with multi-level representations combining global entity and local mention information. We achieve state-of-the-art relation extraction results on the DocRED dataset and report the first entity-level end-to-end relation extraction results for future reference. Finally, our experimental results suggest that a joint approach is on par with task-specific learning, though more efficient due to shared parameters and training steps.