November 22, 2014
Language differences cause geographical and cultural divisions online. Moreover, social media platforms such as Facebook and Twitter seek to optimize exposure to information according to features such as user interests, background, and social context. This raises a concern that users are becoming trapped in their own personalized “filter bubbles,” exposed only to opinions that conform to their beliefs and political positions, potentially creating information “islands” and social polarization. Assistant Professor of Information and Library Science Xiaozhong Liu and his collaborators are developing text mining methods to connect concepts in two different languages on similar social media platforms (English—Twitter and Chinese—Weibo), comparing information shared on the two platforms in terms of topics and networks, in order to quantify the “language bubble” phenomenon.
Professor Liu presented the latest findings of this research at the Rob Kling Center for Social Media Colloquium on November 14, 2014. He and his research team had a paper accepted by the ACM Conference on Hypertext and Social Media earlier this year, and a second paper was recently submitted for consideration to another conference.
Shuai, X., Liu, Xiaozhong, Xia, T., Wu, Y., & Guo, C. (2014). Comparing the pulses of categorical hot events in Twitter and Weibo. Proceedings of the 25th ACM Conference on Hypertext and Social Media, pp. 126-135. New York: ACM.