Generative Adversarial Networks for Annotated Data Augmentation in Data Sparse NLU

TL;DR

This paper demonstrates that using sequential GANs for data augmentation significantly improves NLU model performance in low-resource and new language scenarios, reducing manual annotation efforts.

Contribution

It introduces novel sequential GAN architectures with token and sentence-level rewards for effective data augmentation in NLU tasks.

Findings

Synthetic data boosts NLU accuracy across metrics.

Transfer learning of embeddings enhances GAN performance.

GAN-based augmentation outperforms simple upsampling.

Abstract

Data sparsity is one of the key challenges associated with model development in Natural Language Understanding (NLU) for conversational agents. The challenge is made more complex by the demand for high quality annotated utterances commonly required for supervised learning, usually resulting in weeks of manual labor and high cost. In this paper, we present our results on boosting NLU model performance through training data augmentation using a sequential generative adversarial network (GAN). We explore data generation in the context of two tasks, the bootstrapping of a new language and the handling of low resource features. For both tasks we explore three sequential GAN architectures, one with a token-level reward function, another with our own implementation of a token-level Monte Carlo rollout reward, and a third with sentence-level reward. We evaluate the performance of these feedback models across several sampling methodologies and compare our results to upsampling the original data to the same scale. We further improve the GAN model performance through the transfer learning of the pretrained embeddings. Our experiments reveal synthetic data generated using the sequential generative adversarial network provides significant performance boosts across multiple metrics and can be a major benefit to the NLU tasks.