Abstract: Considering the problem of quantitative evaluation for qualitative survivability data and indexes of network information systems, the proposed method gives a multi-level survivability evaluation framework based on cloud model theory and AHP method. The algorithm builts a multi-level evaluation indexes system and compute the weight of each index using AHP method. Moreover, a quantitative method for qualitative data and a fuzzy evaluation method are designed via multidimensional cloud model. Finally, the experimental results show that the proposed method achieves a satisfactory result in quantitative evaluation of complex qualitative indexes, and has certain advantages in terms of accuracy and fuzzy computing.