Novel Chapter Abstractive Summarization using Spinal Tree Aware Sub-Sentential Content Selection

TL;DR

This paper introduces a spinal tree aware extractive-abstractive summarization method for novel chapters, leveraging syntactic constituent information and margin ranking loss to improve content selection and summary quality.

Contribution

It proposes a novel extractive component utilizing spinal tree structures and margin ranking loss, enhancing the extraction process for long, complex chapter summaries.

Findings

Achieved a 3.71 ROUGE-1 score improvement over prior methods.

Effectively handles lengthy inputs with skewed datasets.

Enhances extractive summarization with syntactic context.

Abstract

Summarizing novel chapters is a difficult task due to the input length and the fact that sentences that appear in the desired summaries draw content from multiple places throughout the chapter. We present a pipelined extractive-abstractive approach where the extractive step filters the content that is passed to the abstractive component. Extremely lengthy input also results in a highly skewed dataset towards negative instances for extractive summarization; we thus adopt a margin ranking loss for extraction to encourage separation between positive and negative examples. Our extraction component operates at the constituent level; our approach to this problem enriches the text with spinal tree information which provides syntactic context (in the form of constituents) to the extraction model. We show an improvement of 3.71 Rouge-1 points over best results reported in prior work on an existing novel chapter dataset.