Improving Relation Extraction with Knowledge-attention

TL;DR

This paper introduces a knowledge-attention encoder that integrates external lexical knowledge into deep neural networks, enhancing relation extraction performance beyond traditional self-attention models.

Contribution

It presents a novel knowledge-attention mechanism and three integration methods with self-attention, achieving state-of-the-art results on TACRED dataset.

Findings

Knowledge-attention complements self-attention effectively.

Integrated models outperform CNN, RNN, and self-attention baselines.

Achieves state-of-the-art on TACRED dataset.

Abstract

While attention mechanisms have been proven to be effective in many NLP tasks, majority of them are data-driven. We propose a novel knowledge-attention encoder which incorporates prior knowledge from external lexical resources into deep neural networks for relation extraction task. Furthermore, we present three effective ways of integrating knowledge-attention with self-attention to maximize the utilization of both knowledge and data. The proposed relation extraction system is end-to-end and fully attention-based. Experiment results show that the proposed knowledge-attention mechanism has complementary strengths with self-attention, and our integrated models outperform existing CNN, RNN, and self-attention based models. State-of-the-art performance is achieved on TACRED, a complex and large-scale relation extraction dataset.