Tracking sentiment and topic dynamics from social media

Yulan He; Chenghua Lin; Wei Gao; Kam-Fai Wong

Tracking sentiment and topic dynamics from social media

Yulan He, Chenghua Lin, Wei Gao, Kam-Fai Wong

Computing Science

Research output: Chapter in Book/Report/Conference proceeding › Published conference contribution

23 Citations (Scopus)

Abstract

We propose a dynamic joint sentiment-topic model (dJST) which allows the detection and tracking of views of current and recurrent interests and shifts in topic and sentiment. Both topic and sentiment dynamics are captured by assuming that the current sentiment-topic specific word distributions are generated according to the word distributions at previous epochs. We derive efficient online inference procedures to sequentially update the model with newly arrived data and show the effectiveness of our proposed model on the Mozilla add-on reviews crawled between 2007 and 2011.

Original language	English
Title of host publication	Proceedings of the 6th International AAAI Conference on Weblogs and Social Media (ICWSM)
Publisher	Association for the Advancement of Artificial Intelligence
Pages	483-486
Number of pages	4
Publication status	Published - 2012

Bibliographical note

Acknowledgements
The authors would like to thank Dong Liu for crawling Mozilla add-ons review data. This work was partially supported by the EC-FP7 project ROBUST (grant number 257859) and the Short Award funded by the Royal Academy of Engineering, UK.

Access to Document

http://www.aaai.org/ocs/index.php/ICWSM/ICWSM12/paper/view/4496

Cite this

@inproceedings{0b1ddea516e8410bb0c211ea3af33239,

title = "Tracking sentiment and topic dynamics from social media",

abstract = "We propose a dynamic joint sentiment-topic model (dJST) which allows the detection and tracking of views of current and recurrent interests and shifts in topic and sentiment. Both topic and sentiment dynamics are captured by assuming that the current sentiment-topic specific word distributions are generated according to the word distributions at previous epochs. We derive efficient online inference procedures to sequentially update the model with newly arrived data and show the effectiveness of our proposed model on the Mozilla add-on reviews crawled between 2007 and 2011.",

author = "Yulan He and Chenghua Lin and Wei Gao and Kam-Fai Wong",

note = "Acknowledgements The authors would like to thank Dong Liu for crawling Mozilla add-ons review data. This work was partially supported by the EC-FP7 project ROBUST (grant number 257859) and the Short Award funded by the Royal Academy of Engineering, UK.",

year = "2012",

language = "English",

pages = "483--486",

booktitle = "Proceedings of the 6th International AAAI Conference on Weblogs and Social Media (ICWSM)",

publisher = "Association for the Advancement of Artificial Intelligence",

}

TY - GEN

T1 - Tracking sentiment and topic dynamics from social media

AU - He, Yulan

AU - Lin, Chenghua

AU - Gao, Wei

AU - Wong, Kam-Fai

N1 - Acknowledgements The authors would like to thank Dong Liu for crawling Mozilla add-ons review data. This work was partially supported by the EC-FP7 project ROBUST (grant number 257859) and the Short Award funded by the Royal Academy of Engineering, UK.

PY - 2012

Y1 - 2012

N2 - We propose a dynamic joint sentiment-topic model (dJST) which allows the detection and tracking of views of current and recurrent interests and shifts in topic and sentiment. Both topic and sentiment dynamics are captured by assuming that the current sentiment-topic specific word distributions are generated according to the word distributions at previous epochs. We derive efficient online inference procedures to sequentially update the model with newly arrived data and show the effectiveness of our proposed model on the Mozilla add-on reviews crawled between 2007 and 2011.

AB - We propose a dynamic joint sentiment-topic model (dJST) which allows the detection and tracking of views of current and recurrent interests and shifts in topic and sentiment. Both topic and sentiment dynamics are captured by assuming that the current sentiment-topic specific word distributions are generated according to the word distributions at previous epochs. We derive efficient online inference procedures to sequentially update the model with newly arrived data and show the effectiveness of our proposed model on the Mozilla add-on reviews crawled between 2007 and 2011.

M3 - Published conference contribution

SP - 483

EP - 486

BT - Proceedings of the 6th International AAAI Conference on Weblogs and Social Media (ICWSM)

PB - Association for the Advancement of Artificial Intelligence

ER -