Health-related spammer detection on chinese social media

Xinhuan Chen, Yong Zhang, Jennifer Xu, Chunxiao Xing, Hsinchun Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Weibo (Chinese microblog) has become a popular social media platform for users to share health-related information. However, illegitimate users or spammers often generate and spread false or misleading health information so as to advertise and attract more attention. To address this issue, we propose a healthrelated spammer detection approach on Chinese social media. Our approach is a deep belief network (DBN) based model incorporating a comprehensive feature set, including burstiness-based features, profile-based features, and content-based features, to identify spammers who spread misleading health-related information. Especially, we create a medical and health domain lexicon to better extract content-based features. The experimental results show the approach achieves an F1 score of 86% in detecting spammer and significantly outperforms the benchmark methods using baseline features.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages284-295
Number of pages12
Volume9545
ISBN (Print)9783319291741
DOIs
StatePublished - 2016
EventInternational Conference for Smart Health, ICSH 2015 - Phoenix, United States
Duration: Nov 17 2015Nov 18 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9545
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

OtherInternational Conference for Smart Health, ICSH 2015
CountryUnited States
CityPhoenix
Period11/17/1511/18/15

Fingerprint

Social Media
Health
Belief Networks
Bayesian networks
Baseline
Benchmark
Experimental Results
Model

Keywords

  • Chinese
  • Deep belief network
  • Health
  • Spammer detection
  • Weibo

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Chen, X., Zhang, Y., Xu, J., Xing, C., & Chen, H. (2016). Health-related spammer detection on chinese social media. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9545, pp. 284-295). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9545). Springer Verlag. https://doi.org/10.1007/978-3-319-29175-8_27

Health-related spammer detection on chinese social media. / Chen, Xinhuan; Zhang, Yong; Xu, Jennifer; Xing, Chunxiao; Chen, Hsinchun.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9545 Springer Verlag, 2016. p. 284-295 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9545).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Chen, X, Zhang, Y, Xu, J, Xing, C & Chen, H 2016, Health-related spammer detection on chinese social media. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9545, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9545, Springer Verlag, pp. 284-295, International Conference for Smart Health, ICSH 2015, Phoenix, United States, 11/17/15. https://doi.org/10.1007/978-3-319-29175-8_27
Chen X, Zhang Y, Xu J, Xing C, Chen H. Health-related spammer detection on chinese social media. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9545. Springer Verlag. 2016. p. 284-295. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-29175-8_27
Chen, Xinhuan ; Zhang, Yong ; Xu, Jennifer ; Xing, Chunxiao ; Chen, Hsinchun. / Health-related spammer detection on chinese social media. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9545 Springer Verlag, 2016. pp. 284-295 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{83ef36bbba1141268cf85f3ab638a0fb,
title = "Health-related spammer detection on chinese social media",
abstract = "Weibo (Chinese microblog) has become a popular social media platform for users to share health-related information. However, illegitimate users or spammers often generate and spread false or misleading health information so as to advertise and attract more attention. To address this issue, we propose a healthrelated spammer detection approach on Chinese social media. Our approach is a deep belief network (DBN) based model incorporating a comprehensive feature set, including burstiness-based features, profile-based features, and content-based features, to identify spammers who spread misleading health-related information. Especially, we create a medical and health domain lexicon to better extract content-based features. The experimental results show the approach achieves an F1 score of 86{\%} in detecting spammer and significantly outperforms the benchmark methods using baseline features.",
keywords = "Chinese, Deep belief network, Health, Spammer detection, Weibo",
author = "Xinhuan Chen and Yong Zhang and Jennifer Xu and Chunxiao Xing and Hsinchun Chen",
year = "2016",
doi = "10.1007/978-3-319-29175-8_27",
language = "English (US)",
isbn = "9783319291741",
volume = "9545",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "284--295",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Health-related spammer detection on chinese social media

AU - Chen, Xinhuan

AU - Zhang, Yong

AU - Xu, Jennifer

AU - Xing, Chunxiao

AU - Chen, Hsinchun

PY - 2016

Y1 - 2016

N2 - Weibo (Chinese microblog) has become a popular social media platform for users to share health-related information. However, illegitimate users or spammers often generate and spread false or misleading health information so as to advertise and attract more attention. To address this issue, we propose a healthrelated spammer detection approach on Chinese social media. Our approach is a deep belief network (DBN) based model incorporating a comprehensive feature set, including burstiness-based features, profile-based features, and content-based features, to identify spammers who spread misleading health-related information. Especially, we create a medical and health domain lexicon to better extract content-based features. The experimental results show the approach achieves an F1 score of 86% in detecting spammer and significantly outperforms the benchmark methods using baseline features.

AB - Weibo (Chinese microblog) has become a popular social media platform for users to share health-related information. However, illegitimate users or spammers often generate and spread false or misleading health information so as to advertise and attract more attention. To address this issue, we propose a healthrelated spammer detection approach on Chinese social media. Our approach is a deep belief network (DBN) based model incorporating a comprehensive feature set, including burstiness-based features, profile-based features, and content-based features, to identify spammers who spread misleading health-related information. Especially, we create a medical and health domain lexicon to better extract content-based features. The experimental results show the approach achieves an F1 score of 86% in detecting spammer and significantly outperforms the benchmark methods using baseline features.

KW - Chinese

KW - Deep belief network

KW - Health

KW - Spammer detection

KW - Weibo

UR - http://www.scopus.com/inward/record.url?scp=84958581024&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958581024&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-29175-8_27

DO - 10.1007/978-3-319-29175-8_27

M3 - Conference contribution

AN - SCOPUS:84958581024

SN - 9783319291741

VL - 9545

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 284

EP - 295

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -