E-commerce abnormal attack detection method

The invention belongs to the technical field of electronic commerce, and particularly relates to an e-commerce abnormal attack detection method. The detection method comprises the following steps: corresponding to e-commerce evaluation parameters according to different flow data characteristic values; and classifying the e-commerce according to the e-commerce evaluation parameters to obtain e-commerce sets, each e-commerce set corresponding to a set of abnormal flow analysis model for evaluating whether the e-commerce evaluation parameter corresponding to a certain e-commerce is within a normal threshold range and analyzing a client type, judging whether an attack is an abnormal attack according to an analysis result, and executing a corresponding protection instruction. Compared with an existing anomaly detection library, deep analysis needs to be conducted on e-commerce information, and simplification is achieved. By further judging whether early warning information can be sent to the e-commerce or freezing