Abstract:ive text summarization is an essential task in natural language processing, which aims to generate concise and condensed summaries retaining the salient information of the input document. Despite the progress of previous work, generating summaries, which are informative, grammatically correct and diverse, remains challenging in practice. In this paper, we present a Plausibility-promoting Generative Adversarial Network for Abstractive Text Summarization with Multi-Task constraint (PGAN-ATSMT), which shows promising performance for generating informative, grammatically correct, and novel summaries. First, PGAN-ATSMT adopts a plausibility-promoting generative adversarial network, which jointly trains a discriminative model D and a generative model G via adversarial learning. The generative model G employs the sequence-to-sequence architecture as its backbone, taking as input the original text and generating a corresponding summary. A novel language model based discriminator D is proposed to distinguish the generated summaries by G from the ground truth summaries without the saturation issue in the previous binary classifier discriminator. The generative model G and the discriminative model D are learned with a minimax two-player game, thus this adversarial process can eventually adjust G to produce high-quality and plausible summaries. Second, we propose two extended regularizations for the generative model G using the multi-task learning, sharing its LSTM encoder and LSTM decoder with text categorization task and syntax annotation task, respectively. The auxiliary tasks help to improve the quality of locating salient information of a document and generate high-quality summaries from language modeling perspective alleviating the issues of incomplete sentences and duplicated words. Experimental results on two benchmark datasets illustrate that PGAN-ATSMT achieves better performance than the state-of-the-art baseline methods in terms of both quantitative and qualitative evaluations.

Adversarial Subsequences for Unconditional Text Generation

Adversarial Sub-sequence for Text Generation.

Sequence Generative Adversarial Nets with a Conditional Discriminator

Adversarial Learning for Neural Dialogue Generation.

Adding A Filter Based on The Discriminator to Improve Unconditional Text Generation

Generative Adversarial Network for Abstractive Text Summarization.

Generating Natural Language Adversarial Examples on a Large Scale with Generative Models

SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient

Adversarial Text Generation Via Sequence Contrast Discrimination.

Improving Conditional Sequence Generative Adversarial Networks by Stepwise Evaluation

Semi-supervised Text Regression with Conditional Generative Adversarial Networks

Generating Long and Coherent Text with Multi-Level Generative Adversarial Networks

Adversarial Feature Matching for Text Generation

Sub-Gan: An Unsupervised Generative Model Via Subspaces

Text Feature Adversarial Learning for Text Generation With Knowledge Transfer From GPT2

Plausibility-promoting Generative Adversarial Network for Abstractive Text Summarization with Multi-Task Constraint.

Long Text Generation via Adversarial Training with Leaked Information

EnsembleGAN: Adversarial Learning for Retrieval-Generation Ensemble Model on Short-Text Conversation

Generating Reasonable and Diversified Story Ending Using Sequence to Sequence Model with Adversarial Training.

A Sequence-to-Sequence&Set Model for Text-to-Table Generation