The invention discloses an anti-communication network fraud identification method based on feature classification, and relates to the technical field of anti-communication network fraud, and the method comprises the following steps: S1, data receiving; S2, data analysis; S3, data feedback confirmation; S4, performing manual auditing processing; S5, processing suspected data; and S6, processing result displaying, wherein, in the step S1, the data acquisition device internally comprises a data acquisition unit, a data processing unit and a CPU interface, the CPU interface is electrically connected with the data acquisition unit, and the data acquisition unit is electrically connected with the data processing unit. According to the method, the network fraud data can be better intercepted andcontrolled, the identification degree of the network fraud data is higher, external personnel are effectively prevented from receiving the network fraud, the unstored fraud data can be transmitted into the big database to be stored, an operator can match the next group of fraud data conveniently, and the identification and interception efficiency is improved.