Quality Function Deployment(QFD) is a new product development process. Its basic concept is to deploy the voice of customer into a final product through product planning, part planning, process planning, and manufacturing. In the product planning stage, which is the first stage of product development, customer requirements or customer attributes(CAs) are translated into engineering characteristics(ECs). Then, based on the relationships between CAs and ECs, the target values of ECs are determined. The engineering characteristic values determined are used as data for making a decision in the subsequent stage. In the previous research, the process of analyzing these relationships is accomplished in a subjective and qualitative manner. In addition, there is no formal methodology for determining the target values of ECs. In this thesis, we formulate the process of determining the target values of ECs as an optimization model. First, we determine the relationships between CAs and ECs as cumulative logit models based on statistical approach, and construct constraints into which company's strategy as well as the needs of customers can be incorporated. Next, cost functions of ECs are developed, which are summed into an objective function. Finally, an algorithm to solve the formulated optimization problem is developed.