Leveraging multi-level dependency of relational sequences for social spammer detection

Abstract Much recent research has shed light on developing the relation-dependent but content-independent framework for social spammer detection. This is mainly because relation among users difficult to be altered when spammers attempt conceal their malicious intentions. Our study investigates detection problem in context of multi-relation networks and makes an fully exploit sequences heterogeneous relations enhancing accuracy. Specifically, we present Multi-level Dependency Model (MDM). The MDM able user’s long-term dependency hidden relational along with short-term dependency. Moreover, considers from perspectives individual-level union-level, due fact that type multi-folds. Experimental results a real-world multi-relational network demonstrate effectiveness our proposed