Topic-Guided Abstractive Text Summarization: a Joint Learning Approach

TL;DR

This paper presents a novel topic-guided abstractive summarization model that integrates neural topic modeling with a Transformer-based seq2seq framework, improving summary quality by capturing global document semantics.

Contribution

It introduces a joint learning approach combining neural topic modeling with Transformer-based summarization, enhancing the capture of global semantics for better summaries.

Findings

Outperforms existing extractive and abstractive models in ROUGE scores

Achieves higher human evaluation scores

Demonstrates effectiveness on multiple datasets

Abstract

We introduce a new approach for abstractive text summarization, Topic-Guided Abstractive Summarization, which calibrates long-range dependencies from topic-level features with globally salient content. The idea is to incorporate neural topic modeling with a Transformer-based sequence-to-sequence (seq2seq) model in a joint learning framework. This design can learn and preserve the global semantics of the document, which can provide additional contextual guidance for capturing important ideas of the document, thereby enhancing the generation of summary. We conduct extensive experiments on two datasets and the results show that our proposed model outperforms many extractive and abstractive systems in terms of both ROUGE measurements and human evaluation. Our code is available at: https://github.com/chz816/tas.