Big data based user clustering and influence power ranking

Telecom big data implies abundant user information. In this paper, it employs the telecom data and proposes a user clustering and influence power ranking scheme. The scheme is implemented through three stages, i.e. the user portrait analysis stage, the user clustering analysis stage and the ranking stage of user influence power. Experimental results have shown that, marketing promotion effectiveness based on this scheme has been improved significantly, while the advertising costs are also considerably reduced.