Abstract:Nowadays, neural text generation has made tremendous progress in abstractive summarization tasks. However, most of the existing summarization models take in the whole document all at once, which sometimes cannot meet the needs in practice. Practically, social text streams such as news events and tweets keep growing from time to time, and can only be fed to the summarization system step by step. Hence, in this paper, we propose the task of Stepwise Summarization, which aims to generate a new appended summary each time a new document is proposed. The appended summary should not only summarize the newly added content but also be coherent with the previous summary, to form an up-to-date complete summary. To tackle this challenge, we design an adversarial learning model, named Stepwise Summary Generator (SSG). First, SSG selectively processes the new document under the guidance of the previous summary, obtaining polished document representation. Next, SSG generates the summary considering both the previous summary and the document. Finally, a convolutional-based discriminator is employed to determine whether the newly generated summary is coherent with the previous summary. For the experiment, we extend the traditional two-step update summarization setting to a multi-step stepwise setting, and re-propose a large-scale stepwise summarization dataset based on a public story generation dataset. Extensive experiments on this dataset show that SSG achieves state-of-the-art performance in terms of both automatic metrics and human evaluations. Ablation studies demonstrate the effectiveness of each module in our framework. We also discuss the benefits and limitations of recent large language models on this task.

Semantic-Preserving Abstractive Text Summarization with Siamese Generative Adversarial Net.

A GAN Based Video Summarization Method with Representation Loss

Automatic Document Summarization Via Deep Neural Networks

A Syntax-Augmented and Headline-Aware Neural Text Summarization Method

Text Summarization Generation Based on Semantic Similarity

SemSUM: Semantic Dependency Guided Neural Abstractive Summarization

Generative Adversarial Network for Abstractive Text Summarization.

Topic-Aware Abstractive Text Summarization

Abstract Summarization Model Based on Semantic Graphs and Entity Pointers

Abstractive Summarization Improved by WordNet-based Extractive Sentences

Abstractive method-based Text Summarization using Bidirectional Long Short-Term Memory and Pointer Generator Mode

Learning to Encode Text As Human-Readable Summaries Using Generative Adversarial Networks.

An Abstractive Summarizer Based on Improved Pointer-Generator Network

Joint Parsing and Generation for Abstractive Summarization

Topic-Guided Abstractive Text Summarization: a Joint Learning Approach

Abstractive Document Summarization with a Graph-Based Attentional Neural Model.

Write Summary Step-by-Step: A Pilot Study of Stepwise Summarization

Abstractive Text Summarization Using GAN

Abstractive Text Summarization by Incorporating Reader Comments.

A Semantic Relevance Based Neural Network for Text Summarization and Text Simplification

GATSum: Graph-Based Topic-Aware Abstract Text Summarization