The evaluation of DDos based on ERS is a process that recognizes the level type of a sample by comparing it with the standard values from the data in the ERS index. However, it is difficult to find the same value as those standard values in practice. As the above mentioned procedure is a typical fuzzy pattern recognition problem, a fuzzy pattern recognition methodology is proposed to solve it. The standard value matrix of factors is constructed on the basis of the data in the ERS index and then converted into the membership degree matrix. A new fuzzy pattern recognition model is developed to provide the evaluation levels of samples, which are expressed by linguistic variables.