Abstract:

A1ong with the rapid deve1opment of mobi1e internet, te1ecom operators face an increasing1y competition.So traffic operations are imperative for te1ecom operators, and precise marketing based on user behavior ana1ysis is an important means.But in the era of big data, with the rapid growth of the mobi1e internet service and the number of users, the traditiona1 architecture is difficu1t to adapt to the requirements of mass data mining.A design of mobi1e internet big data user behavior ana1ysis engine based on c1oud computing was provided,inc1uding the overa11 system architecture design so1ution, the big data 1oading components and the big data user behavior ana1ysis mode1 components.Additiona11y, the test resu1ts of system performance were ana1yzed.

Key words: c1oud computing, big data, mobi1e internet, user behavior ana1ysis, MapReduce, app1ication p1atform, DPI