Improving Multi-Document Summarization via Text Classification

TL;DR

This paper introduces TCSum, a novel multi-document summarization system that leverages text classification data and representations to improve performance and generate style-specific summaries, overcoming data scarcity issues.

Contribution

The paper presents TCSum, a new summarization approach that uses text classification data and distributed representations to enhance multi-document summarization without handcrafted features.

Findings

Achieves state-of-the-art results on DUC datasets

Effectively captures summary style variations across categories

Does not rely on handcrafted features

Abstract

Developed so far, multi-document summarization has reached its bottleneck due to the lack of sufficient training data and diverse categories of documents. Text classification just makes up for these deficiencies. In this paper, we propose a novel summarization system called TCSum, which leverages plentiful text classification data to improve the performance of multi-document summarization. TCSum projects documents onto distributed representations which act as a bridge between text classification and summarization. It also utilizes the classification results to produce summaries of different styles. Extensive experiments on DUC generic multi-document summarization datasets show that, TCSum can achieve the state-of-the-art performance without using any hand-crafted features and has the capability to catch the variations of summary styles with respect to different text categories.