Bringing Structure into Summaries: a Faceted Summarization Dataset for Long Scientific Documents (2021-05-31T00:00:00.000000Z)

TL;DR

Faceted summarization will spur further advances in summarization research and foster the development of NLP systems that can leverage the structured information in both long texts and summaries, according to this study.

Abstract

Faceted summarization provides briefings of a document from different perspectives. Readers can quickly comprehend the main points of a long document with the help of a structured outline. However, little research has been conducted on this subject, partially due to the lack of large-scale faceted summarization datasets. In this study, we present FacetSum, a faceted summarization benchmark built on Emerald journal articles, covering a diverse range of domains. Different from traditional document-summary pairs, FacetSum provides multiple summaries, each targeted at specific sections of a long document, including the purpose, method, findings, and value. Analyses and empirical results on our dataset reveal the importance of bringing structure into summaries. We believe FacetSum will spur further advances in summarization research and foster the development of NLP systems that can leverage the structured information in both long texts and summaries.

Authors

Xingdi Yuan

11 papers

Rui Meng

3 papers

Tong Wang

3 papers

TL;DR

Abstract

Authors

References41 items

Multi-XScience: A Large-scale Dataset for Extreme Multi-document Summarization of Scientific Articles

SEAL: Segment-wise Extractive-Abstractive Long-form Text Summarization

CODA-19: Using a Non-Expert Crowd to Annotate Research Aspects on 10,000+ Abstracts in the COVID-19 Open Research Dataset

HipoRank: Incorporating Hierarchical and Positional Information into Graph-based Unsupervised Long Document Extractive Summarization

TLDR: Extreme Summarization of Scientific Documents

A Divide-and-Conquer Approach to the Summarization of Long Documents

BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension

ScisummNet: A Large Annotated Corpus and Content-Impact Models for Scientific Paper Summarization with Citation Networks

BIGPATENT: A Large-Scale Dataset for Abstractive and Coherent Summarization

Structured Summarization of Academic Publications

fairseq: A Fast, Extensible Toolkit for Sequence Modeling

Insights from CL-SciSumm 2016: the faceted scientific document summarization Shared Task

Don’t Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization

Beyond Generic Summarization: A Multi-faceted Hierarchical Summarization Corpus of Large Heterogeneous Data

Newsroom: A Dataset of 1.3 Million Summaries with Diverse Extractive Strategies

A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents

A Supervised Approach to Extractive Summarisation of Scientific Papers

A Deep Reinforced Model for Abstractive Summarization

Get To The Point: Summarization with Pointer-Generator Networks

SummaRuNNer: A Recurrent Neural Network Based Sequence Model for Extractive Summarization of Documents

Current findings from research on structured abstracts: an update.

Using Argumentative Zones for Extractive Summarization of Scientific Articles

Automatic classification of sentences to support Evidence Based Medicine

Beyond SumBasic: Task-focused summarization with sentence simplification and lexical expansion

LexRank: Graph-based Lexical Centrality as Salience in Text Summarization

TextRank: Bringing Order into Text

Generic text summarization using relevance measure and latent semantic analysis

How to write a paper for publication.

Are Structured Abstracts Easier to Read Than Traditional Ones

Obtaining information accurately and quickly: are structured abstracts more efficient?

Structure of the scientific paper.

sumy: Automatic text summarizer

Writing an article abstract

On Extractive and Abstractive Neural Document Summarization with Transformer Language Models

Implementation Details To make BART take full text as input, we extend the positional embedding to 10,000 tokens

spaCy: Industrial-strength Natural Language Processing in Python

Emerald 110k: A Multidisciplinary Dataset for Abstract Sentence Classification

ScisummNet: A Large Annotated Corpus and Content-Impact Models for Scientiﬁc Paper Summarization with Citation Networks

Systematic generalization: What is required and can it be learned?

Overview of the CL-SciSumm 2016 Shared Task

Full results (Rouge-1/2/L) of different models on FacetSum. Bold text indicates the best scores on

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names