Decision-making is one of the important human's characteristic and every body makes so much decision daily. As personal and social life could be affected by correct and well-timed decision-making, existence of scientific methods that helps every one in this field is necessary. In ranked voting systems, each voter selects a subset of the candidates and ranks them in order of preference. For this reason, several methods to discriminate among efficient candidates have been proposed. Each method has some weaknesses and strengths. Some of these methods aren’t stable and monotonic, such as Nogchi's model. In this paper, Noguchi's model is improved through changing constraints. Because of constraints corresponding to inefficient candidates are redundant, a test removed them before using the model. This test is based on a transformation on the given data which makes it possible to use the definition of efficiency instead of AR-efficiency.Then it's became stable and monotonic. Furthermore, the algorithm was proposed that can rank candidates with votes of voters and managers’ opinion. In other part of this research, a nonlinear programming model was proposed that could rank candidates with weighted kashida; TEXT-ALIGN: justify; LINE-HEIGHT: 70%; TEXT-KASHIDA: 0%; MARGIN: 0cm 0cm 0pt" Key words Dta Envelopment Analysis, Preference Voting System, Noguchi's Model, Manager's Opinion, Voter's Weighted ltr"