Abstract: There are many factors influencing the stability of tailings dam，which are coupled with each other and have a lot of unascertained information.In order to solve this problem，the stability of tailings dam is considered as the result factor，and 9 qualitative factors and 18 quantitative factors influencing the result factor are selected as condition factors to construct the factor space for attribute recognition of tailings dam stability.The unascertained mathematics theory is introduced into the factor space to construct the factor space-unascertained measure model， and the solution method of this model is discussed.The subjective and objective weights are calculated by analytic hierarchy process (AHP) and information entropy theory，and the synthetical weight of each condition factor is determined.By using factor synthesis method，the high dimension of factor space is reduced to one dimension combined factor axis，and the identification vector of combined factor is obtained.The stability of a tailings dam is evaluated according to the confidence criterion.This model is applied to the engineering practice to verify its feasibility and applicability.The results show that the attribute recognition results of tailings dam stability is basically consistent with the field discrimination results，which indicated thet the model established in this paper is help for solving the tailings dam stability evaluation problem and could provide valuable reference for diagnosis and evaluation of similar complex systems.

Key words: Tailings dam, Attribute recognition of tailings dam stability, Factor space, Unascertained measure, Combination weight, Dimensionality reduction