Detecting Users with Multiple Aliases on Twitter

In day to day life, many people create their multiple identities on social networking sites in order to abstract their true identity. No wonder, monitoring and analysis of the social networks are becoming important for analysts since terrorists and extremists are using the information provided by the user, to spread advocacy and communicating with each other. As Twitter is a widely used social network to share opinions and views, in this work, we have applied Supervised Machine Learning (i.e. alias detection) on the Twitter dataset. Here, we propose a technique for detecting users who make use of aliases in which time profiles are incorporated with the stylometric-based features. By combining time profiles and stylometric features, the detection rate of a user having multiple aliases increases significantly. Further, we make use of data mining techniques to improve accuracy. The obtained results show that time profiles can pose to be very useful tools for alias detection when combined with stylometry.