Detection of zombie followers in SINA Weibo
Zhedi Zhang, Futai Zou, Pan Li, Bei Pei, Jianhua Li · 2016
SINA Weibo is one of the most popular social networks in China where people can share everything in the form of pictures, texts, videos or links. The prime motivation pushing people to post tweet is to increase their number of followers. However, some artificial followers that can be bought and sold online, called zombie followers, emerge in endlessly. This paper introduces a series of algorithms to detect the zombie followers efficiently and accurately by analyzing two kinds of users' features. Firstly distinguish the zombie followers out of the legitimate users manually, and then utilize SVM machine learning classifier to detect the zombie followers automatically. Finally, through the estimation on the real dataset, the zombie followers can be correctly detected out of the legitimate users at high accuracy (99.78%) and the false positive rate is 11.57%.