The mobile broadband business currently is in early development. In broadband telecommunications

Unicom broadband and other third-party broadband operators have fixed the size of the user base. The goal of the development of mobile broadband optical should be focused in the use of mobile customers

accurately identify mobile customers suspected of different network users. Currently

different broadband network recognition using artificial judgment

the recognition accurate rate is low and the number of users covered is too little. This paper through the analysis of mining method analysis of the features of the use of different types of broadband users

with the accurate identification of different network broadband users

achieve suspected different network broadband users list output

for the operations staff to carry out different network broadband users turn network marketing activities and expand operator broadband market share to provide customer information support.