Abstract

Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.

Original languageEnglish (US)
Pages (from-to)1-25
Number of pages25
JournalNew Review of Hypermedia and Multimedia
DOIs
StateAccepted/In press - Jun 26 2018

Fingerprint

Factorization
Electric network analysis
Experiments
Polarization

Keywords

  • dynamic graph algorithms
  • Negative link
  • non-negative matrix tri-factorization
  • online political networks
  • online social networks

ASJC Scopus subject areas

  • Information Systems
  • Media Technology
  • Computer Science Applications

Cite this

Implicit negative link detection on online political networks via matrix tri-factorizations. / Ozer, Mert; Yildirim, Mehmet Yigit; Davulcu, Hasan.

In: New Review of Hypermedia and Multimedia, 26.06.2018, p. 1-25.

Research output: Contribution to journalArticle

@article{025a6517d9334f24a893c7610f339e46,
title = "Implicit negative link detection on online political networks via matrix tri-factorizations",
abstract = "Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.",
keywords = "dynamic graph algorithms, Negative link, non-negative matrix tri-factorization, online political networks, online social networks",
author = "Mert Ozer and Yildirim, {Mehmet Yigit} and Hasan Davulcu",
year = "2018",
month = "6",
day = "26",
doi = "10.1080/13614568.2018.1482964",
language = "English (US)",
pages = "1--25",
journal = "New Review of Hypermedia and Multimedia",
issn = "1361-4568",
publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Implicit negative link detection on online political networks via matrix tri-factorizations

AU - Ozer, Mert

AU - Yildirim, Mehmet Yigit

AU - Davulcu, Hasan

PY - 2018/6/26

Y1 - 2018/6/26

N2 - Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.

AB - Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.

KW - dynamic graph algorithms

KW - Negative link

KW - non-negative matrix tri-factorization

KW - online political networks

KW - online social networks

UR - http://www.scopus.com/inward/record.url?scp=85049020285&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049020285&partnerID=8YFLogxK

U2 - 10.1080/13614568.2018.1482964

DO - 10.1080/13614568.2018.1482964

M3 - Article

SP - 1

EP - 25

JO - New Review of Hypermedia and Multimedia

JF - New Review of Hypermedia and Multimedia

SN - 1361-4568

ER -