Predicting Personal Information of Facebook Users Based on What Fan Pages They Like

• Main Authors: Siang-Min Wang, 王湘閔
• Other Authors: 林敏勝
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/ye5hd7

碩士 === 國立臺北科技大學 === 電機工程系所 === 105 === With the Internet development and becoming more widespread, online social networks gained more popularity. Facebook is the most influential social networking site in the world. Fan page is one of the most popular services provided by Facebook. Nowadays, more than 50 million fan pages set up on Facebook. This thesis applies data mining techniques to predict and analyze the users personal information based on what fan pages they like on Facebook. The experimental results show that (1) the gender prediction on the dataset with birthday attribute is more accurate than that without birthday attribute; (2) the age prediction on female dataset is more accurate than that on male dataset; (3) the most liked fan pages on Facebook for females with 24-31 years old are fashion and beauty related.