Personalized knowledge visualization in twitter

Abstract

In recent years, Twitter has been playing an essential role in broadcasting real-time news and useful information. Many commercial media companies have used Twitter as an effective online tool to broadcast breaking news. Consequently, the tweets are organized in chronological order. In this paper, we investigate the tweet exhibition problem from a new perspective. We treat Twitter as an important source of knowledge gaining and exploit its semantic attributes. In particular, we focus on extracting and distilling knowledge from tweets and proposing a new organization and visualization tool to exhibit the tweets in a semantic manner. We conducted our experiments on Amazon Turk and the user feedback from crowd demonstrated the effectiveness of our new Twitter visualization tool.