JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]


Journal of Information Science and Engineering, Vol. 34 No. 6, pp. 1579-1598


SocialDNA: A Novel Approach for Distinguishing Notable Articles and Authors through Social Events


MING-HUNG WANG1 AND CHIN-LAUNG LEI2
1Department of Information Engineering and Computer Science
Feng Chia University
Taichung, 407 Taiwan

2Department of Electrical Engineering
National Taiwan University
Taipei, 106 Taiwan
E-mail: mhwang@fcu.edu.tw; cllei@ntu.edu.tw


With the rapid development of online social networks, increasing amounts of usergenerated content are posted online. While information is overwhelming a site, readers actually prefer fully reading some influential articles or information rather than glancing sequentially at every article on a forum. In this paper, we propose techniques for forecasting public responses to articles shortly after the articles are published. Our proposal identifies the important articles from two perspectives: frequent discussion and extreme acceptance or rejection by the online public. We also discuss approaches for distinguishing influential authors who are popular and receive consistently high ratings from online users. To verify our methodologies, we analyzed a popular social forum in Taiwan during three large-scale social events, a social movement and two national election campaigns. Our results demonstrate that our methodologies achieve high accuracy with significant time reduction and outperform previous methodologies in distinguishing notable articles and authors on social forums.


Keywords: social networks, social media, information filtering, opinion leader, public opinion

  Retrieve PDF document (JISE_201806_12.pdf)