The index weight determines the scientificity and rationality of the comprehensive evaluation results. However, the traditional index calculation method, which is scored directly by decision-makers, excessively depends on the degree of experience and knowledge accumulation of decision-makers. Different decision-makers’ understandings of the index have great differences and fuzziness, which makes the weight calculation results biased or even wrong due to the influence of human subjective factors. To solve this problem, this paper considers that the index weight value contains the decision-makers’ personal subjective information and common objective information, and then proposes a weight calculation approach to improve the objectivity of the index based on the Superiority of Neighboring Objectives and the Ensemble Empirical Mode Decomposition, which can eliminate the subjective information in the weight value and obtain a more objective weight value. Finally, the rationality and objectivity of the proposed method are verified by comparing them with the classical AHP, subjective weighting method and objective weighting method.