Abstract

The rapid growth of social media services brings a large amount of high-dimensional social media data at an unprecedented rate. Feature selection is powerful to prepare high-dimensional data by finding a subset of relevant features. A vast majority of existing feature selection algorithms for social media data exclusively focus on positive interactions among linked instances such as friendships and user following relations. However, in many real-world social networks, instances may also be negatively interconnected. Recent work shows that negative links have an added value over positive links in advancing many learning tasks. In this paper, we study a novel problem of unsupervised feature selection in signed social networks and propose a novel framework SignedFS. In particular, we provide a principled way to model positive and negative links for user latent representation learning. Then we embed the user latent representations into feature selection when label information is not available. Also, we revisit the principle of homophily and balance theory in signed social networks and incorporate the signed graph regularization into the feature selection framework to capture the first-order and the second-order proximity among users in signed social networks. Experiments on two real-world signed social networks demonstrate the effectiveness of our proposed framework. Further experiments are conducted to understand the impacts of different components of SignedFS.

Original languageEnglish (US)
Title of host publicationKDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PublisherAssociation for Computing Machinery
Pages777-786
Number of pages10
VolumePart F129685
ISBN (Electronic)9781450348874
DOIs
StatePublished - Aug 13 2017
Event23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017 - Halifax, Canada
Duration: Aug 13 2017Aug 17 2017

Other

Other23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017
CountryCanada
CityHalifax
Period8/13/178/17/17

Fingerprint

Feature extraction
Labels
Experiments

Keywords

  • Feature selection
  • Signed social networks
  • Unsupervised learning

ASJC Scopus subject areas

  • Software
  • Information Systems

Cite this

Cheng, K., Li, J., & Liu, H. (2017). Unsupervised feature selection in signed social networks. In KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Vol. Part F129685, pp. 777-786). Association for Computing Machinery. https://doi.org/10.1145/3097983.3098106

Unsupervised feature selection in signed social networks. / Cheng, Kewei; Li, Jundong; Liu, Huan.

KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685 Association for Computing Machinery, 2017. p. 777-786.

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

Cheng, K, Li, J & Liu, H 2017, Unsupervised feature selection in signed social networks. in KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. vol. Part F129685, Association for Computing Machinery, pp. 777-786, 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017, Halifax, Canada, 8/13/17. https://doi.org/10.1145/3097983.3098106
Cheng K, Li J, Liu H. Unsupervised feature selection in signed social networks. In KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685. Association for Computing Machinery. 2017. p. 777-786 https://doi.org/10.1145/3097983.3098106
Cheng, Kewei ; Li, Jundong ; Liu, Huan. / Unsupervised feature selection in signed social networks. KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Vol. Part F129685 Association for Computing Machinery, 2017. pp. 777-786
@inproceedings{177bbdcc6dbd4fbaa5f4bb9c5791d32b,
title = "Unsupervised feature selection in signed social networks",
abstract = "The rapid growth of social media services brings a large amount of high-dimensional social media data at an unprecedented rate. Feature selection is powerful to prepare high-dimensional data by finding a subset of relevant features. A vast majority of existing feature selection algorithms for social media data exclusively focus on positive interactions among linked instances such as friendships and user following relations. However, in many real-world social networks, instances may also be negatively interconnected. Recent work shows that negative links have an added value over positive links in advancing many learning tasks. In this paper, we study a novel problem of unsupervised feature selection in signed social networks and propose a novel framework SignedFS. In particular, we provide a principled way to model positive and negative links for user latent representation learning. Then we embed the user latent representations into feature selection when label information is not available. Also, we revisit the principle of homophily and balance theory in signed social networks and incorporate the signed graph regularization into the feature selection framework to capture the first-order and the second-order proximity among users in signed social networks. Experiments on two real-world signed social networks demonstrate the effectiveness of our proposed framework. Further experiments are conducted to understand the impacts of different components of SignedFS.",
keywords = "Feature selection, Signed social networks, Unsupervised learning",
author = "Kewei Cheng and Jundong Li and Huan Liu",
year = "2017",
month = "8",
day = "13",
doi = "10.1145/3097983.3098106",
language = "English (US)",
volume = "Part F129685",
pages = "777--786",
booktitle = "KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - Unsupervised feature selection in signed social networks

AU - Cheng, Kewei

AU - Li, Jundong

AU - Liu, Huan

PY - 2017/8/13

Y1 - 2017/8/13

N2 - The rapid growth of social media services brings a large amount of high-dimensional social media data at an unprecedented rate. Feature selection is powerful to prepare high-dimensional data by finding a subset of relevant features. A vast majority of existing feature selection algorithms for social media data exclusively focus on positive interactions among linked instances such as friendships and user following relations. However, in many real-world social networks, instances may also be negatively interconnected. Recent work shows that negative links have an added value over positive links in advancing many learning tasks. In this paper, we study a novel problem of unsupervised feature selection in signed social networks and propose a novel framework SignedFS. In particular, we provide a principled way to model positive and negative links for user latent representation learning. Then we embed the user latent representations into feature selection when label information is not available. Also, we revisit the principle of homophily and balance theory in signed social networks and incorporate the signed graph regularization into the feature selection framework to capture the first-order and the second-order proximity among users in signed social networks. Experiments on two real-world signed social networks demonstrate the effectiveness of our proposed framework. Further experiments are conducted to understand the impacts of different components of SignedFS.

AB - The rapid growth of social media services brings a large amount of high-dimensional social media data at an unprecedented rate. Feature selection is powerful to prepare high-dimensional data by finding a subset of relevant features. A vast majority of existing feature selection algorithms for social media data exclusively focus on positive interactions among linked instances such as friendships and user following relations. However, in many real-world social networks, instances may also be negatively interconnected. Recent work shows that negative links have an added value over positive links in advancing many learning tasks. In this paper, we study a novel problem of unsupervised feature selection in signed social networks and propose a novel framework SignedFS. In particular, we provide a principled way to model positive and negative links for user latent representation learning. Then we embed the user latent representations into feature selection when label information is not available. Also, we revisit the principle of homophily and balance theory in signed social networks and incorporate the signed graph regularization into the feature selection framework to capture the first-order and the second-order proximity among users in signed social networks. Experiments on two real-world signed social networks demonstrate the effectiveness of our proposed framework. Further experiments are conducted to understand the impacts of different components of SignedFS.

KW - Feature selection

KW - Signed social networks

KW - Unsupervised learning

UR - http://www.scopus.com/inward/record.url?scp=85029037412&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029037412&partnerID=8YFLogxK

U2 - 10.1145/3097983.3098106

DO - 10.1145/3097983.3098106

M3 - Conference contribution

AN - SCOPUS:85029037412

VL - Part F129685

SP - 777

EP - 786

BT - KDD 2017 - Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

PB - Association for Computing Machinery

ER -