User network data fingerprint constructing method and system based on distributed processing and DPI data

The invention discloses a user network data fingerprint constructing method and system based on a distributed processing frame and DPI data so as to extract online behavior characteristics and references of a user mobile terminal. The system comprises a data preprocessing module for cleaning and eliminating redundancy of original data for meeting the requirement for network data fingerprints, a rule extraction module, a user set extracting module, a user behavior extracting module and a data storage module, wherein the rule extracting module selects M common mobile phone APPs, obtains the domain name of each APP through packet capture and performs regex matching, and the matching type is adopted as recognition rules of each APP and forms rule files; the user set extracting module extractsuser sets needing to be counted by the network data fingerprint system; the user behavior extracting module counts the access condition of M APPs of the user within each unit period of time; the datastorage module stores resu