Finding the influential Users using Text Mining on Twitter

• Main Authors: Cheng-DaoLee, 李政道
• Other Authors: Yau-Hwang Kuo
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/33168810879410218965

碩士 === 國立成功大學 === 醫學資訊研究所 === 101 === With the development of technology and the consideration of business cost and change of pace of people, more and more people use social network like twitter, facebook, myspace…etc. Social marketing has replaced traditional viral marketing and becoming a hot research issue. What is different from the past is that information-transferring on social marketing is not based on random users but finding the influential users, it could make sure the speed up the information-transferring and efficiency attain the best result. In the thesis, we use the concept of social marketing to propose a user influential ability ranking model to improve the speed of information-transferring and efficiency. At first, we search the relational tweets with specific key terms to analyze the topics and sentiment score. And we use the relationship among tweets and preference degree to create an information flow map. Finally we analyze user profile on twitter to test the different influential models scores and choosing the suitable model to expect in line with actual situation.