Abstract
This paper extends a fuzzy version of Analytic Hierarchy Process to reflect human's vagueness or subjective judgment. New comprehensive technique, variable axis method [VAM], is developed to decide the total ordering of fuzzy numbers. Further, we adopt the fuzzy optimization method in group decision-making, minimizing the total sum of the dissatisfaction of each evaluator. Geometric mean technique is employed to synthesize a fuzzy positive reciprocal matrix. Numerical example demonstrates the effectiveness and the applicability of the proposed decision making model.