Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media

Qiaozhen Guo; Wei (Wayne) Huang; Kai Huang; Xiao Liu

doi:10.1007/978-3-319-29175-8_12

Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media

Qiaozhen Guo, Wei (Wayne) Huang, Kai Huang, Xiao Liu

Information Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

6 Scopus citations

Abstract

Social media is an important data source to compliment traditional epidemic surveillance. However, misinformation in social media hinders the exploitation of valuable information. Analysis of information credibility has drawn much attention of academia in recent years. In this paper, we focus on analyzing the credibility of influenza posts published on Sina Weibo. We propose a semi-supervised probabilistic graphical model to jointly learn the interactions between user trustworthiness, content reliability, and post credibility. To test the performance of the approach, we apply it to identify credible influenza posts published from May 2013 to June 2014 on Sina Weibo. Random Forests and the Bayesian Network are used as baselines for evaluation. The results show that our approach performs effectively with the highest average accuracy of 71.7%, f-measure 51%. Our proposed framework significantly outperformed the baselines in detecting credible influenza posts on Sina Weibo.

Original language	English (US)
Title of host publication	Smart Health - International Conference, ICSH 2015, Revised Selected Papers
Editors	Hsinchun Chen, Daniel Dajun Zeng, Xiaolong Zheng, Scott J. Leischow
Publisher	Springer Verlag
Pages	131-142
Number of pages	12
ISBN (Print)	9783319291741
DOIs	https://doi.org/10.1007/978-3-319-29175-8_12
State	Published - 2016
Event	International Conference for Smart Health, ICSH 2015 - Phoenix, United States Duration: Nov 17 2015 → Nov 18 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9545
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	International Conference for Smart Health, ICSH 2015
Country/Territory	United States
City	Phoenix
Period	11/17/15 → 11/18/15

Keywords

Credibility
Influenza
Objectivity
Probabilistic graphical models
Reliability
Sina weibo
Trustworthiness

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-29175-8_12

Cite this

Guo, Q., Huang, W., Huang, K., & Liu, X. (2016). Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media. In H. Chen, D. D. Zeng, X. Zheng, & S. J. Leischow (Eds.), Smart Health - International Conference, ICSH 2015, Revised Selected Papers (pp. 131-142). (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_12

Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media. / Guo, Qiaozhen; Huang, Wei (Wayne); Huang, Kai et al.
Smart Health - International Conference, ICSH 2015, Revised Selected Papers. ed. / Hsinchun Chen; Daniel Dajun Zeng; Xiaolong Zheng; Scott J. Leischow. Springer Verlag, 2016. p. 131-142 (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 proceeding › Conference contribution

Guo, Q, Huang, W, Huang, K & Liu, X 2016, Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media. in H Chen, DD Zeng, X Zheng & SJ Leischow (eds), Smart Health - International Conference, ICSH 2015, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9545, Springer Verlag, pp. 131-142, International Conference for Smart Health, ICSH 2015, Phoenix, United States, 11/17/15. https://doi.org/10.1007/978-3-319-29175-8_12

Guo Q, Huang W, Huang K, Liu X. Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media. In Chen H, Zeng DD, Zheng X, Leischow SJ, editors, Smart Health - International Conference, ICSH 2015, Revised Selected Papers. Springer Verlag. 2016. p. 131-142. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-29175-8_12

Guo, Qiaozhen ; Huang, Wei (Wayne) ; Huang, Kai et al. / Information credibility : A probabilistic graphical model for identifying credible influenza posts on social media. Smart Health - International Conference, ICSH 2015, Revised Selected Papers. editor / Hsinchun Chen ; Daniel Dajun Zeng ; Xiaolong Zheng ; Scott J. Leischow. Springer Verlag, 2016. pp. 131-142 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{a37cac2a1bcf4985b6cdf2f9abd047c0,

title = "Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media",

abstract = "Social media is an important data source to compliment traditional epidemic surveillance. However, misinformation in social media hinders the exploitation of valuable information. Analysis of information credibility has drawn much attention of academia in recent years. In this paper, we focus on analyzing the credibility of influenza posts published on Sina Weibo. We propose a semi-supervised probabilistic graphical model to jointly learn the interactions between user trustworthiness, content reliability, and post credibility. To test the performance of the approach, we apply it to identify credible influenza posts published from May 2013 to June 2014 on Sina Weibo. Random Forests and the Bayesian Network are used as baselines for evaluation. The results show that our approach performs effectively with the highest average accuracy of 71.7%, f-measure 51%. Our proposed framework significantly outperformed the baselines in detecting credible influenza posts on Sina Weibo.",

keywords = "Credibility, Influenza, Objectivity, Probabilistic graphical models, Reliability, Sina weibo, Trustworthiness",

author = "Qiaozhen Guo and Huang, {Wei (Wayne)} and Kai Huang and Xiao Liu",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2016.; International Conference for Smart Health, ICSH 2015 ; Conference date: 17-11-2015 Through 18-11-2015",

year = "2016",

doi = "10.1007/978-3-319-29175-8_12",

language = "English (US)",

isbn = "9783319291741",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "131--142",

editor = "Hsinchun Chen and Zeng, {Daniel Dajun} and Xiaolong Zheng and Leischow, {Scott J.}",

booktitle = "Smart Health - International Conference, ICSH 2015, Revised Selected Papers",

}

TY - GEN

T1 - Information credibility

T2 - International Conference for Smart Health, ICSH 2015

AU - Guo, Qiaozhen

AU - Huang, Wei (Wayne)

AU - Huang, Kai

AU - Liu, Xiao

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2016.

PY - 2016

Y1 - 2016

N2 - Social media is an important data source to compliment traditional epidemic surveillance. However, misinformation in social media hinders the exploitation of valuable information. Analysis of information credibility has drawn much attention of academia in recent years. In this paper, we focus on analyzing the credibility of influenza posts published on Sina Weibo. We propose a semi-supervised probabilistic graphical model to jointly learn the interactions between user trustworthiness, content reliability, and post credibility. To test the performance of the approach, we apply it to identify credible influenza posts published from May 2013 to June 2014 on Sina Weibo. Random Forests and the Bayesian Network are used as baselines for evaluation. The results show that our approach performs effectively with the highest average accuracy of 71.7%, f-measure 51%. Our proposed framework significantly outperformed the baselines in detecting credible influenza posts on Sina Weibo.

AB - Social media is an important data source to compliment traditional epidemic surveillance. However, misinformation in social media hinders the exploitation of valuable information. Analysis of information credibility has drawn much attention of academia in recent years. In this paper, we focus on analyzing the credibility of influenza posts published on Sina Weibo. We propose a semi-supervised probabilistic graphical model to jointly learn the interactions between user trustworthiness, content reliability, and post credibility. To test the performance of the approach, we apply it to identify credible influenza posts published from May 2013 to June 2014 on Sina Weibo. Random Forests and the Bayesian Network are used as baselines for evaluation. The results show that our approach performs effectively with the highest average accuracy of 71.7%, f-measure 51%. Our proposed framework significantly outperformed the baselines in detecting credible influenza posts on Sina Weibo.

KW - Credibility

KW - Influenza

KW - Objectivity

KW - Probabilistic graphical models

KW - Reliability

KW - Sina weibo

KW - Trustworthiness

UR - http://www.scopus.com/inward/record.url?scp=84958568554&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958568554&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-29175-8_12

DO - 10.1007/978-3-319-29175-8_12

M3 - Conference contribution

AN - SCOPUS:84958568554

SN - 9783319291741

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

SP - 131

EP - 142

BT - Smart Health - International Conference, ICSH 2015, Revised Selected Papers

A2 - Chen, Hsinchun

A2 - Zeng, Daniel Dajun

A2 - Zheng, Xiaolong

A2 - Leischow, Scott J.

PB - Springer Verlag

Y2 - 17 November 2015 through 18 November 2015

ER -

Information credibility: A probabilistic graphical model for identifying credible influenza posts on social media

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this