Twitter Topic Extraction using Principal Component Analysis
Murphy Choy · Research Journal of Social Science & Management · 2012
Twitter has become one of the most popular social network tools among internet users. The micro-blogging tool has enable users to express their opinion online as well as the ability to disseminate the information to other users rapidly. There are several approaches of topic extraction that were proposed which focuses on the use of clustering techniques and page-rank algorithm. However, all the techniques discussed did not attempt to measure the amount of information covered by the algorithm and how well does the topics reflect the actual twitter that they are derived from. In this paper, we will describe the effectiveness of the Principal Component Analysis in extracting the topics from the twitter data and describing the amount of information covered by the topics as a percentage of all the topics in the twitters.