CatGAN: Category-aware Generative Adversarial Networks with Hierarchical Evolutionary Learning for Category Text Generation

TL;DR

This paper introduces CatGAN, a category-aware GAN with hierarchical evolutionary learning, designed to improve multi-category text generation by enhancing quality and diversity while stabilizing training.

Contribution

The paper presents a novel CatGAN model with a hierarchical evolutionary training algorithm and Gumbel-Softmax relaxation for effective multi-category text generation.

Findings

CatGAN outperforms most existing state-of-the-art methods.

Hierarchical evolutionary learning stabilizes training and balances quality and diversity.

Gumbel-Softmax relaxation simplifies training on discrete data.

Abstract

Generating multiple categories of texts is a challenging task and draws more and more attention. Since generative adversarial nets (GANs) have shown competitive results on general text generation, they are extended for category text generation in some previous works. However, the complicated model structures and learning strategies limit their performance and exacerbate the training instability. This paper proposes a category-aware GAN (CatGAN) which consists of an efficient category-aware model for category text generation and a hierarchical evolutionary learning algorithm for training our model. The category-aware model directly measures the gap between real samples and generated samples on each category, then reducing this gap will guide the model to generate high-quality category samples. The Gumbel-Softmax relaxation further frees our model from complicated learning strategies for updating CatGAN on discrete data. Moreover, only focusing on the sample quality normally leads the mode collapse problem, thus a hierarchical evolutionary learning algorithm is introduced to stabilize the training procedure and obtain the trade-off between quality and diversity while training CatGAN. Experimental results demonstrate that CatGAN outperforms most of the existing state-of-the-art methods.