Abstract

The study of collective behavior is to understand how individuals behave in a social network environment. Oceans of data generated by social media like Facebook, Twitter, Flickr and YouTube present opportunities and challenges to studying collective behavior in a large scale. In this work, we aim to learn to predict collective behavior in social media. In particular, given information about some individuals, how can we infer the behavior of unobserved individuals in the same network? A social-dimension based approach is adopted to address the heterogeneity of connections presented in social media. However, the networks in social media are normally of colossal size, involving hundreds of thousands or even millions of actors. The scale of networks entails scalable learning of models for collective behavior prediction. To address the scalability issue, we propose an edge-centric clustering scheme to extract sparse social dimensions. With sparse social dimensions, the social-dimension based approach can efficiently handle networks of millions of actors while demonstrating comparable prediction performance as other non-scalable methods.

Original languageEnglish (US)
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
Pages1107-1116
Number of pages10
DOIs
StatePublished - 2009
EventACM 18th International Conference on Information and Knowledge Management, CIKM 2009 - Hong Kong, China
Duration: Nov 2 2009Nov 6 2009

Other

OtherACM 18th International Conference on Information and Knowledge Management, CIKM 2009
CountryChina
CityHong Kong
Period11/2/0911/6/09

Fingerprint

Social dimension
Social media
Prediction
Clustering
Facebook
Scalability
Network environment
Social networks
Twitter

Keywords

  • Behavior prediction
  • Edge-centric clustering
  • Relational learning
  • Social dimensions
  • Social media

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Decision Sciences(all)

Cite this

Tang, L., & Liu, H. (2009). Scalable learning of collective behavior based on sparse social dimensions. In International Conference on Information and Knowledge Management, Proceedings (pp. 1107-1116) https://doi.org/10.1145/1645953.1646094

Scalable learning of collective behavior based on sparse social dimensions. / Tang, Lei; Liu, Huan.

International Conference on Information and Knowledge Management, Proceedings. 2009. p. 1107-1116.

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

Tang, L & Liu, H 2009, Scalable learning of collective behavior based on sparse social dimensions. in International Conference on Information and Knowledge Management, Proceedings. pp. 1107-1116, ACM 18th International Conference on Information and Knowledge Management, CIKM 2009, Hong Kong, China, 11/2/09. https://doi.org/10.1145/1645953.1646094
Tang L, Liu H. Scalable learning of collective behavior based on sparse social dimensions. In International Conference on Information and Knowledge Management, Proceedings. 2009. p. 1107-1116 https://doi.org/10.1145/1645953.1646094
Tang, Lei ; Liu, Huan. / Scalable learning of collective behavior based on sparse social dimensions. International Conference on Information and Knowledge Management, Proceedings. 2009. pp. 1107-1116
@inproceedings{04dbffa66bad405ab0655acbb240919c,
title = "Scalable learning of collective behavior based on sparse social dimensions",
abstract = "The study of collective behavior is to understand how individuals behave in a social network environment. Oceans of data generated by social media like Facebook, Twitter, Flickr and YouTube present opportunities and challenges to studying collective behavior in a large scale. In this work, we aim to learn to predict collective behavior in social media. In particular, given information about some individuals, how can we infer the behavior of unobserved individuals in the same network? A social-dimension based approach is adopted to address the heterogeneity of connections presented in social media. However, the networks in social media are normally of colossal size, involving hundreds of thousands or even millions of actors. The scale of networks entails scalable learning of models for collective behavior prediction. To address the scalability issue, we propose an edge-centric clustering scheme to extract sparse social dimensions. With sparse social dimensions, the social-dimension based approach can efficiently handle networks of millions of actors while demonstrating comparable prediction performance as other non-scalable methods.",
keywords = "Behavior prediction, Edge-centric clustering, Relational learning, Social dimensions, Social media",
author = "Lei Tang and Huan Liu",
year = "2009",
doi = "10.1145/1645953.1646094",
language = "English (US)",
isbn = "9781605585123",
pages = "1107--1116",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",

}

TY - GEN

T1 - Scalable learning of collective behavior based on sparse social dimensions

AU - Tang, Lei

AU - Liu, Huan

PY - 2009

Y1 - 2009

N2 - The study of collective behavior is to understand how individuals behave in a social network environment. Oceans of data generated by social media like Facebook, Twitter, Flickr and YouTube present opportunities and challenges to studying collective behavior in a large scale. In this work, we aim to learn to predict collective behavior in social media. In particular, given information about some individuals, how can we infer the behavior of unobserved individuals in the same network? A social-dimension based approach is adopted to address the heterogeneity of connections presented in social media. However, the networks in social media are normally of colossal size, involving hundreds of thousands or even millions of actors. The scale of networks entails scalable learning of models for collective behavior prediction. To address the scalability issue, we propose an edge-centric clustering scheme to extract sparse social dimensions. With sparse social dimensions, the social-dimension based approach can efficiently handle networks of millions of actors while demonstrating comparable prediction performance as other non-scalable methods.

AB - The study of collective behavior is to understand how individuals behave in a social network environment. Oceans of data generated by social media like Facebook, Twitter, Flickr and YouTube present opportunities and challenges to studying collective behavior in a large scale. In this work, we aim to learn to predict collective behavior in social media. In particular, given information about some individuals, how can we infer the behavior of unobserved individuals in the same network? A social-dimension based approach is adopted to address the heterogeneity of connections presented in social media. However, the networks in social media are normally of colossal size, involving hundreds of thousands or even millions of actors. The scale of networks entails scalable learning of models for collective behavior prediction. To address the scalability issue, we propose an edge-centric clustering scheme to extract sparse social dimensions. With sparse social dimensions, the social-dimension based approach can efficiently handle networks of millions of actors while demonstrating comparable prediction performance as other non-scalable methods.

KW - Behavior prediction

KW - Edge-centric clustering

KW - Relational learning

KW - Social dimensions

KW - Social media

UR - http://www.scopus.com/inward/record.url?scp=74549120273&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=74549120273&partnerID=8YFLogxK

U2 - 10.1145/1645953.1646094

DO - 10.1145/1645953.1646094

M3 - Conference contribution

AN - SCOPUS:74549120273

SN - 9781605585123

SP - 1107

EP - 1116

BT - International Conference on Information and Knowledge Management, Proceedings

ER -