Visual Analysis of Topic Competition on Social Media

IEEE Transactions on Visualization and Computer Graphics (IEEE VAST 2013)

Panpan Xu*1     Yingcai Wu2     Enxun Wei*1,3     Tai-Quan Peng4   
Shixia Liu2    Jonathan Z.H. Zhu5    Huamin Qu1
Authors associated with * are/were the interns under the supervision of Yingcai Wu in MSRA
1Hong Kong University of Science and Technology      2Microsoft Research Asia
3Shanghai Jiao Tong University      4Nanyang Technological University      5City University of Hong Kong

Teaser Image

The grassroots and the media played the major roles in recruiting public attention to the topic welfare / society on Twitter. The trending keywords related to the topic were not the same at different times. Around June 5th, there was a trending discussion on solyndra, while around June 12th, the keywords "teachers" and "firefighters" gained more significance. The sparklines underlying the keywords have their peaks on the selected time interval, indicating potential correlation between those topics to the recruitment effects observed. The tweets containing the keywords can also be examined to find the actual events that trigger the discussions.

Abstract

How do various topics compete for public attention when they are spreading on social media? What roles do opinion leaders play in the rise and fall of competitiveness of various topics? In this study, we propose an expanded topic competition model to characterize the competition for public attention on multiple topics promoted by various opinion leaders on social media. To allow an intuitive understanding of the estimated measures, we present a timeline visualization through a metaphoric interpretation of the results. The visual design features both topical and social aspects of the information diffusion process by compositing ThemeRiver with storyline style visualization. ThemeRiver shows the increase and decrease of competitiveness of each topic. Opinion leaders are drawn as threads that converge or diverge with regard to their roles in influencing the public agenda change over time. To validate the effectiveness of the visual analysis techniques, we report the insights gained on two collections of Tweets: the 2012 United States presidential election and the Occupy Wall Street movement.

Paper

Download the paper

Video

Download the video (42MB)

Slide Deck
Visual Analysis of Topic Competition on Social Media (slides made by Panpan Xu) from Yingcai Wu

Download the slides (3.3MB)

BibTeX

@article {YWu2013d,
author = {Panpan Xu and Yingcai Wu and Enxun Wei and Tai-Quan Peng and Shixia Liu and Jonathan J.H. Zhu and Huamin Qu},
title = {Visual Analysis of Topic Competition on Social Media},
journal = {IEEE Transactions on Visualization and Computer Graphics (Proceedings of IEEE VAST 2013},
year = {2013},
volume = {19},
number = {12},
pages = {2012--2021}
}

Acknowledgements

We would like to thank Professor Lu Wei from Zhejiang University for suggesting the chance of interdisciplinary collaboration, the anonymous eviewers for their valuable and constructive comments, and Fangzhao Wu for his help. This work is partially supported by a grant from MSRA and RGC GRF 618313.