A3Net: Adversarial-and-Attention Network for Machine Reading Comprehension

TL;DR

A3Net is a novel machine reading comprehension model that combines adversarial training on multiple target variables with a multi-layer attention mechanism, leading to improved robustness and state-of-the-art performance on WebQA.

Contribution

The paper introduces a3Net, integrating adversarial training on multiple targets and a multi-layer attention network for enhanced comprehension.

Findings

Outperforms state-of-the-art models on WebQA with a Fuzzy Score of 77.0%.

Adversarial training improves model robustness and generalization.

Multi-layer attention enhances question-passage interaction.

Abstract

In this paper, we introduce Adversarial-and-attention Network (A3Net) for Machine Reading Comprehension. This model extends existing approaches from two perspectives. First, adversarial training is applied to several target variables within the model, rather than only to the inputs or embeddings. We control the norm of adversarial perturbations according to the norm of original target variables, so that we can jointly add perturbations to several target variables during training. As an effective regularization method, adversarial training improves robustness and generalization of our model. Second, we propose a multi-layer attention network utilizing three kinds of high-efficiency attention mechanisms. Multi-layer attention conducts interaction between question and passage within each layer, which contributes to reasonable representation and understanding of the model. Combining these two contributions, we enhance the diversity of dataset and the information extracting ability of the model at the same time. Meanwhile, we construct A3Net for the WebQA dataset. Results show that our model outperforms the state-of-the-art models (improving Fuzzy Score from 73.50% to 77.0%).