Timely video popularity forecasting based on social networks

Jie Xu, Mihaela Van Der Schaar, Jiangchuan Liu, Haitao Li

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

5 Citations (Scopus)

Abstract

This paper presents Pop-Forecast, a systematic method for accurately forecasting the popularity of videos promoted through social networks. Pop-Forecast aims to optimize the forecasting accuracy and the timeliness with which forecasts are issued, by explicitly taking into account the dynamic propagation of videos in social networks. The forecasting is performed online and requires no training phase or a priori knowledge. We analytically bound the performance loss of Pop-Forecast as compared to that obtained by an omniscient oracle and prove that the bound is sublinear in the number of video arrivals, thereby guaranteeing its fast rate of convergence as well as its asymptotic convergence to the optimal performance. We validate the performance of Pop-Forecast through extensive experiments using real-world data traces collected from the videos shared in RenRen, one of the largest online social networks in China. These experiments show that our proposed method outperforms existing approaches for popularity prediction (which do not take into account the propagation in social network) by more than 30% in terms of prediction rewards.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE INFOCOM
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2308-2316
Number of pages9
Volume26
ISBN (Print)9781479983810
DOIs
StatePublished - Aug 21 2015
Externally publishedYes
Event34th IEEE Annual Conference on Computer Communications and Networks, IEEE INFOCOM 2015 - Hong Kong, Hong Kong
Duration: Apr 26 2015May 1 2015

Other

Other34th IEEE Annual Conference on Computer Communications and Networks, IEEE INFOCOM 2015
CountryHong Kong
CityHong Kong
Period4/26/155/1/15

Fingerprint

Experiments

ASJC Scopus subject areas

  • Computer Science(all)
  • Electrical and Electronic Engineering

Cite this

Xu, J., Van Der Schaar, M., Liu, J., & Li, H. (2015). Timely video popularity forecasting based on social networks. In Proceedings - IEEE INFOCOM (Vol. 26, pp. 2308-2316). [7218618] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/INFOCOM.2015.7218618

Timely video popularity forecasting based on social networks. / Xu, Jie; Van Der Schaar, Mihaela; Liu, Jiangchuan; Li, Haitao.

Proceedings - IEEE INFOCOM. Vol. 26 Institute of Electrical and Electronics Engineers Inc., 2015. p. 2308-2316 7218618.

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

Xu, J, Van Der Schaar, M, Liu, J & Li, H 2015, Timely video popularity forecasting based on social networks. in Proceedings - IEEE INFOCOM. vol. 26, 7218618, Institute of Electrical and Electronics Engineers Inc., pp. 2308-2316, 34th IEEE Annual Conference on Computer Communications and Networks, IEEE INFOCOM 2015, Hong Kong, Hong Kong, 4/26/15. https://doi.org/10.1109/INFOCOM.2015.7218618
Xu J, Van Der Schaar M, Liu J, Li H. Timely video popularity forecasting based on social networks. In Proceedings - IEEE INFOCOM. Vol. 26. Institute of Electrical and Electronics Engineers Inc. 2015. p. 2308-2316. 7218618 https://doi.org/10.1109/INFOCOM.2015.7218618
Xu, Jie ; Van Der Schaar, Mihaela ; Liu, Jiangchuan ; Li, Haitao. / Timely video popularity forecasting based on social networks. Proceedings - IEEE INFOCOM. Vol. 26 Institute of Electrical and Electronics Engineers Inc., 2015. pp. 2308-2316
@inproceedings{c3e0a46728704dbcadef3da5913435e3,
title = "Timely video popularity forecasting based on social networks",
abstract = "This paper presents Pop-Forecast, a systematic method for accurately forecasting the popularity of videos promoted through social networks. Pop-Forecast aims to optimize the forecasting accuracy and the timeliness with which forecasts are issued, by explicitly taking into account the dynamic propagation of videos in social networks. The forecasting is performed online and requires no training phase or a priori knowledge. We analytically bound the performance loss of Pop-Forecast as compared to that obtained by an omniscient oracle and prove that the bound is sublinear in the number of video arrivals, thereby guaranteeing its fast rate of convergence as well as its asymptotic convergence to the optimal performance. We validate the performance of Pop-Forecast through extensive experiments using real-world data traces collected from the videos shared in RenRen, one of the largest online social networks in China. These experiments show that our proposed method outperforms existing approaches for popularity prediction (which do not take into account the propagation in social network) by more than 30{\%} in terms of prediction rewards.",
author = "Jie Xu and {Van Der Schaar}, Mihaela and Jiangchuan Liu and Haitao Li",
year = "2015",
month = "8",
day = "21",
doi = "10.1109/INFOCOM.2015.7218618",
language = "English (US)",
isbn = "9781479983810",
volume = "26",
pages = "2308--2316",
booktitle = "Proceedings - IEEE INFOCOM",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Timely video popularity forecasting based on social networks

AU - Xu, Jie

AU - Van Der Schaar, Mihaela

AU - Liu, Jiangchuan

AU - Li, Haitao

PY - 2015/8/21

Y1 - 2015/8/21

N2 - This paper presents Pop-Forecast, a systematic method for accurately forecasting the popularity of videos promoted through social networks. Pop-Forecast aims to optimize the forecasting accuracy and the timeliness with which forecasts are issued, by explicitly taking into account the dynamic propagation of videos in social networks. The forecasting is performed online and requires no training phase or a priori knowledge. We analytically bound the performance loss of Pop-Forecast as compared to that obtained by an omniscient oracle and prove that the bound is sublinear in the number of video arrivals, thereby guaranteeing its fast rate of convergence as well as its asymptotic convergence to the optimal performance. We validate the performance of Pop-Forecast through extensive experiments using real-world data traces collected from the videos shared in RenRen, one of the largest online social networks in China. These experiments show that our proposed method outperforms existing approaches for popularity prediction (which do not take into account the propagation in social network) by more than 30% in terms of prediction rewards.

AB - This paper presents Pop-Forecast, a systematic method for accurately forecasting the popularity of videos promoted through social networks. Pop-Forecast aims to optimize the forecasting accuracy and the timeliness with which forecasts are issued, by explicitly taking into account the dynamic propagation of videos in social networks. The forecasting is performed online and requires no training phase or a priori knowledge. We analytically bound the performance loss of Pop-Forecast as compared to that obtained by an omniscient oracle and prove that the bound is sublinear in the number of video arrivals, thereby guaranteeing its fast rate of convergence as well as its asymptotic convergence to the optimal performance. We validate the performance of Pop-Forecast through extensive experiments using real-world data traces collected from the videos shared in RenRen, one of the largest online social networks in China. These experiments show that our proposed method outperforms existing approaches for popularity prediction (which do not take into account the propagation in social network) by more than 30% in terms of prediction rewards.

UR - http://www.scopus.com/inward/record.url?scp=84954230168&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84954230168&partnerID=8YFLogxK

U2 - 10.1109/INFOCOM.2015.7218618

DO - 10.1109/INFOCOM.2015.7218618

M3 - Conference contribution

AN - SCOPUS:84954230168

SN - 9781479983810

VL - 26

SP - 2308

EP - 2316

BT - Proceedings - IEEE INFOCOM

PB - Institute of Electrical and Electronics Engineers Inc.

ER -