Popularity prediction of social multimedia based on concept drift

Shih Hong Jheng, Cheng-Te Li, Hsi Lin Chen, Man Kwan Shan

研究成果: Conference contribution

2 引文 (Scopus)

摘要

Microblogging services such as Twitter and Plurk allow users to easily access and share different types of social multimedia (e.g. images and videos) over the online social world. However, information overload happens to users and prohibits them from reaching popular and important digital contents. This paper studies the problem of predicting the popularity of social multimedia which is embedded in short messages of microblogging social networks. Social multimedia exhibits the property that they might be persistently or periodically re-shared and thus their popularity might resurrect at some time and evolve over time. We exploit the idea of concept drift to capture this property. We formulate the problem using classification, and propose to tackle the tasks of Re-share classification and Popularity Score classification. Two categories of features are devised and extracted, including information diffusion and explicit multimedia meta information. We develop a concept drift-based popularity predictor, by ensembling multiple trained classifiers from social multimedia instances in different time intervals. The key lies in dynamically determining the ensemble weights of classifiers. Experiments conducted on the Plurk data show the high accuracy on the popularity classification and the promising results on detecting popular social multimedia.

原文English
主出版物標題Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013
頁面821-826
頁數6
DOIs
出版狀態Published - 2013 十二月 1
事件2013 ASE/IEEE Int. Conf. on Social Computing, SocialCom 2013, the 2013 ASE/IEEE Int. Conf. on Big Data, BigData 2013, the 2013 Int. Conf. on Economic Computing, EconCom 2013, the 2013 PASSAT 2013, and the 2013 ASE/IEEE Int. Conf. on BioMedCom 2013 - Washington, DC, United States
持續時間: 2013 九月 82013 九月 14

出版系列

名字Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013

Other

Other2013 ASE/IEEE Int. Conf. on Social Computing, SocialCom 2013, the 2013 ASE/IEEE Int. Conf. on Big Data, BigData 2013, the 2013 Int. Conf. on Economic Computing, EconCom 2013, the 2013 PASSAT 2013, and the 2013 ASE/IEEE Int. Conf. on BioMedCom 2013
國家United States
城市Washington, DC
期間13-09-0813-09-14

指紋

Classifiers
Experiments

All Science Journal Classification (ASJC) codes

  • Software

引用此文

Jheng, S. H., Li, C-T., Chen, H. L., & Shan, M. K. (2013). Popularity prediction of social multimedia based on concept drift. 於 Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013 (頁 821-826). [6693420] (Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013). https://doi.org/10.1109/SocialCom.2013.123
Jheng, Shih Hong ; Li, Cheng-Te ; Chen, Hsi Lin ; Shan, Man Kwan. / Popularity prediction of social multimedia based on concept drift. Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013. 2013. 頁 821-826 (Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013).
@inproceedings{d272c8afd5ce435ebba77a1ece9e6b40,
title = "Popularity prediction of social multimedia based on concept drift",
abstract = "Microblogging services such as Twitter and Plurk allow users to easily access and share different types of social multimedia (e.g. images and videos) over the online social world. However, information overload happens to users and prohibits them from reaching popular and important digital contents. This paper studies the problem of predicting the popularity of social multimedia which is embedded in short messages of microblogging social networks. Social multimedia exhibits the property that they might be persistently or periodically re-shared and thus their popularity might resurrect at some time and evolve over time. We exploit the idea of concept drift to capture this property. We formulate the problem using classification, and propose to tackle the tasks of Re-share classification and Popularity Score classification. Two categories of features are devised and extracted, including information diffusion and explicit multimedia meta information. We develop a concept drift-based popularity predictor, by ensembling multiple trained classifiers from social multimedia instances in different time intervals. The key lies in dynamically determining the ensemble weights of classifiers. Experiments conducted on the Plurk data show the high accuracy on the popularity classification and the promising results on detecting popular social multimedia.",
author = "Jheng, {Shih Hong} and Cheng-Te Li and Chen, {Hsi Lin} and Shan, {Man Kwan}",
year = "2013",
month = "12",
day = "1",
doi = "10.1109/SocialCom.2013.123",
language = "English",
isbn = "9780769551371",
series = "Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013",
pages = "821--826",
booktitle = "Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013",

}

Jheng, SH, Li, C-T, Chen, HL & Shan, MK 2013, Popularity prediction of social multimedia based on concept drift. 於 Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013., 6693420, Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013, 頁 821-826, 2013 ASE/IEEE Int. Conf. on Social Computing, SocialCom 2013, the 2013 ASE/IEEE Int. Conf. on Big Data, BigData 2013, the 2013 Int. Conf. on Economic Computing, EconCom 2013, the 2013 PASSAT 2013, and the 2013 ASE/IEEE Int. Conf. on BioMedCom 2013, Washington, DC, United States, 13-09-08. https://doi.org/10.1109/SocialCom.2013.123

Popularity prediction of social multimedia based on concept drift. / Jheng, Shih Hong; Li, Cheng-Te; Chen, Hsi Lin; Shan, Man Kwan.

Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013. 2013. p. 821-826 6693420 (Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013).

研究成果: Conference contribution

TY - GEN

T1 - Popularity prediction of social multimedia based on concept drift

AU - Jheng, Shih Hong

AU - Li, Cheng-Te

AU - Chen, Hsi Lin

AU - Shan, Man Kwan

PY - 2013/12/1

Y1 - 2013/12/1

N2 - Microblogging services such as Twitter and Plurk allow users to easily access and share different types of social multimedia (e.g. images and videos) over the online social world. However, information overload happens to users and prohibits them from reaching popular and important digital contents. This paper studies the problem of predicting the popularity of social multimedia which is embedded in short messages of microblogging social networks. Social multimedia exhibits the property that they might be persistently or periodically re-shared and thus their popularity might resurrect at some time and evolve over time. We exploit the idea of concept drift to capture this property. We formulate the problem using classification, and propose to tackle the tasks of Re-share classification and Popularity Score classification. Two categories of features are devised and extracted, including information diffusion and explicit multimedia meta information. We develop a concept drift-based popularity predictor, by ensembling multiple trained classifiers from social multimedia instances in different time intervals. The key lies in dynamically determining the ensemble weights of classifiers. Experiments conducted on the Plurk data show the high accuracy on the popularity classification and the promising results on detecting popular social multimedia.

AB - Microblogging services such as Twitter and Plurk allow users to easily access and share different types of social multimedia (e.g. images and videos) over the online social world. However, information overload happens to users and prohibits them from reaching popular and important digital contents. This paper studies the problem of predicting the popularity of social multimedia which is embedded in short messages of microblogging social networks. Social multimedia exhibits the property that they might be persistently or periodically re-shared and thus their popularity might resurrect at some time and evolve over time. We exploit the idea of concept drift to capture this property. We formulate the problem using classification, and propose to tackle the tasks of Re-share classification and Popularity Score classification. Two categories of features are devised and extracted, including information diffusion and explicit multimedia meta information. We develop a concept drift-based popularity predictor, by ensembling multiple trained classifiers from social multimedia instances in different time intervals. The key lies in dynamically determining the ensemble weights of classifiers. Experiments conducted on the Plurk data show the high accuracy on the popularity classification and the promising results on detecting popular social multimedia.

UR - http://www.scopus.com/inward/record.url?scp=84893630126&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893630126&partnerID=8YFLogxK

U2 - 10.1109/SocialCom.2013.123

DO - 10.1109/SocialCom.2013.123

M3 - Conference contribution

AN - SCOPUS:84893630126

SN - 9780769551371

T3 - Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013

SP - 821

EP - 826

BT - Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013

ER -

Jheng SH, Li C-T, Chen HL, Shan MK. Popularity prediction of social multimedia based on concept drift. 於 Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013. 2013. p. 821-826. 6693420. (Proceedings - SocialCom/PASSAT/BigData/EconCom/BioMedCom 2013). https://doi.org/10.1109/SocialCom.2013.123