Linguistic ?? numbers, as well as fuzzy entropy for linguistic ?? numbers are defined. For the multi-criteria decision-making problems that the criteria values are linguistic ?? numbers with incomplete certain information on weights of criteria, a method based on fuzzy entropy and evidential reasoning is proposed. In this method, the criteria weights are attained by constructing a linear programming model based on the linguistic ?? numbers fuzzy entropy, and the evidential
reasoning approach is used to reach the comprehensive criteria values, and the most desirable alternatives are obtained. Analysis of an example demonstrates the feasibility and effectiveness of the method.