Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization

Zhongyu Wei, Wei Gao

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Single-document summarization is a challenging task. In this chapter, we explore effective ways using the tweets linking to news for generating extractive summary of each document. We reveal the very basic value of tweets that can be utilized by regarding every tweet as a vote for candidate sentences. Base on such finding, we resort to unsupervised summarization models by leveraging the linking tweets to master the ranking of candidate extracts via random walk on a heterogeneous graph. The advantage is that we can use the linking tweets to opportunistically “supervise” the summarization with no need of reference summaries. Furthermore, we analyze the influence of the volume and latency of tweets on the quality of output summaries since tweets come after news release. Compared to truly supervised summarizer unaware of tweets, our method achieves significantly better results with reasonably small tradeoff on latency; compared to the same using tweets as auxiliary features, our method is comparable while needing less tweets and much shorter time to achieve significant outperformance.

Original languageEnglish
Title of host publicationSocial Media Content Analysis
Subtitle of host publicationNatural Language Processing and Beyond
PublisherWorld Scientific Publishing Co. Pte Ltd
Pages297-307
Number of pages11
ISBN (Electronic)9789813223615
ISBN (Print)9789813223608
DOIs
Publication statusPublished - 1 Jan 2017

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Wei, Z., & Gao, W. (2017). Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization. In Social Media Content Analysis: Natural Language Processing and Beyond (pp. 297-307). World Scientific Publishing Co. Pte Ltd. https://doi.org/10.1142/9789813223615_0020

Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization. / Wei, Zhongyu; Gao, Wei.

Social Media Content Analysis: Natural Language Processing and Beyond. World Scientific Publishing Co. Pte Ltd, 2017. p. 297-307.

Research output: Chapter in Book/Report/Conference proceedingChapter

Wei, Z & Gao, W 2017, Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization. in Social Media Content Analysis: Natural Language Processing and Beyond. World Scientific Publishing Co. Pte Ltd, pp. 297-307. https://doi.org/10.1142/9789813223615_0020
Wei Z, Gao W. Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization. In Social Media Content Analysis: Natural Language Processing and Beyond. World Scientific Publishing Co. Pte Ltd. 2017. p. 297-307 https://doi.org/10.1142/9789813223615_0020
Wei, Zhongyu ; Gao, Wei. / Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization. Social Media Content Analysis: Natural Language Processing and Beyond. World Scientific Publishing Co. Pte Ltd, 2017. pp. 297-307
@inbook{cdb4577be0a34bf5b55c474f29f000cc,
title = "Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization",
abstract = "Single-document summarization is a challenging task. In this chapter, we explore effective ways using the tweets linking to news for generating extractive summary of each document. We reveal the very basic value of tweets that can be utilized by regarding every tweet as a vote for candidate sentences. Base on such finding, we resort to unsupervised summarization models by leveraging the linking tweets to master the ranking of candidate extracts via random walk on a heterogeneous graph. The advantage is that we can use the linking tweets to opportunistically “supervise” the summarization with no need of reference summaries. Furthermore, we analyze the influence of the volume and latency of tweets on the quality of output summaries since tweets come after news release. Compared to truly supervised summarizer unaware of tweets, our method achieves significantly better results with reasonably small tradeoff on latency; compared to the same using tweets as auxiliary features, our method is comparable while needing less tweets and much shorter time to achieve significant outperformance.",
author = "Zhongyu Wei and Wei Gao",
year = "2017",
month = "1",
day = "1",
doi = "10.1142/9789813223615_0020",
language = "English",
isbn = "9789813223608",
pages = "297--307",
booktitle = "Social Media Content Analysis",
publisher = "World Scientific Publishing Co. Pte Ltd",
address = "Singapore",

}

TY - CHAP

T1 - Gibberish, assistant, or master? Using tweets linking to news for extractive single-document summarization

AU - Wei, Zhongyu

AU - Gao, Wei

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Single-document summarization is a challenging task. In this chapter, we explore effective ways using the tweets linking to news for generating extractive summary of each document. We reveal the very basic value of tweets that can be utilized by regarding every tweet as a vote for candidate sentences. Base on such finding, we resort to unsupervised summarization models by leveraging the linking tweets to master the ranking of candidate extracts via random walk on a heterogeneous graph. The advantage is that we can use the linking tweets to opportunistically “supervise” the summarization with no need of reference summaries. Furthermore, we analyze the influence of the volume and latency of tweets on the quality of output summaries since tweets come after news release. Compared to truly supervised summarizer unaware of tweets, our method achieves significantly better results with reasonably small tradeoff on latency; compared to the same using tweets as auxiliary features, our method is comparable while needing less tweets and much shorter time to achieve significant outperformance.

AB - Single-document summarization is a challenging task. In this chapter, we explore effective ways using the tweets linking to news for generating extractive summary of each document. We reveal the very basic value of tweets that can be utilized by regarding every tweet as a vote for candidate sentences. Base on such finding, we resort to unsupervised summarization models by leveraging the linking tweets to master the ranking of candidate extracts via random walk on a heterogeneous graph. The advantage is that we can use the linking tweets to opportunistically “supervise” the summarization with no need of reference summaries. Furthermore, we analyze the influence of the volume and latency of tweets on the quality of output summaries since tweets come after news release. Compared to truly supervised summarizer unaware of tweets, our method achieves significantly better results with reasonably small tradeoff on latency; compared to the same using tweets as auxiliary features, our method is comparable while needing less tweets and much shorter time to achieve significant outperformance.

UR - http://www.scopus.com/inward/record.url?scp=85041586820&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041586820&partnerID=8YFLogxK

U2 - 10.1142/9789813223615_0020

DO - 10.1142/9789813223615_0020

M3 - Chapter

SN - 9789813223608

SP - 297

EP - 307

BT - Social Media Content Analysis

PB - World Scientific Publishing Co. Pte Ltd

ER -