Abstract:Automatic summarization plays an important role in the exponential document growth on the Web. On content websites such as <a class="link-external link-http" href="http://CNN.com" rel="external noopener nofollow">this http URL</a> and <a class="link-external link-http" href="http://WikiHow.com" rel="external noopener nofollow">this http URL</a>, there often exist various kinds of side information along with the main document for attention attraction and easier understanding, such as videos, images, and queries. Such information can be used for better summarization, as they often explicitly or implicitly mention the essence of the article. However, most of the existing side-aware summarization methods are designed to incorporate either single-modal or multi-modal side information, and cannot effectively adapt to each other. In this paper, we propose a general summarization framework, which can flexibly incorporate various modalities of side information. The main challenges in designing a flexible summarization model with side information include: (1) the side information can be in textual or visual format, and the model needs to align and unify it with the document into the same semantic space, (2) the side inputs can contain information from various aspects, and the model should recognize the aspects useful for summarization. To address these two challenges, we first propose a unified topic encoder, which jointly discovers latent topics from the document and various kinds of side information. The learned topics flexibly bridge and guide the information flow between multiple inputs in a graph encoder through a topic-aware interaction. We secondly propose a triplet contrastive learning mechanism to align the single-modal or multi-modal information into a unified semantic space, where the summary quality is enhanced by better understanding the document and side information. Results show that our model significantly surpasses strong baselines on three public single-modal or multi-modal benchmark summarization datasets.

Using Relevant Public Posts to Enhance News Article Summarization.

Utilizing Microblogs for Automatic News Highlights Extraction

Using Tweets To Help Sentence Compression For News Highlights Generation

Atsum: Extracting Attractive Summaries for News Propagation on Microblogs

News Event Summarization Complemented by Micropoints.

Comparative News Summarization Using Concept-Based Optimization

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

Comparative News Summarization Using Linear Programming

Effective Approaches to Combining Lexical and Syntactical Information for Code Summarization

An Exploration of Post-Editing Effectiveness in Text Summarization

Automatic Text Summarization for Public Health WeChat Official Accounts Platform Base on Improved TextRank

Topic-Focused Summarization Of News Events Based On Biased Snippet Extraction And Selection

End-to-End Segmentation-based News Summarization

Chinese news text opinion summarization based on integrating sentences opinion and topic similarity

Summarizing Event Opinions for News Reports

A Topic-aware Summarization Framework with Different Modal Side Information

Summarizing Events from Massive News Reports on the Web

A New Method for Extractive Text Summarization Using Neural Networks

THE RESEARCH ON SENTENCE FEATURE AND FUZZY INFERENCE FOR CHINESE BREAKING NEWS SUMMARIZATION

Topic-based automatic summarization algorithm for Chinese short text

What is in the news on a subject : automatic and sparse summarization of large document corpora