品質機能展開是一項新產品設計規劃的系統化分析工具。不同於現行的研究以數值或模糊數作為各項評估資訊的表達方式，本研究利用二元模糊語意表達模式作為品質屋各項評估資訊的表達方式，以改善資訊遺漏的問題。在品質屋的建構中，本研究從顧客與潛在顧客，利用德菲法(Delphi method)以二元模糊語意表達模式評估各項顧客需求的重要性、賣點與價值等三項指標。考慮重要性、賣點與價值等三項指標具有不同的重要性權重，本研究利用順序權重幾何平均(ordered weighted geometric averaging, OWGA)運算子整合上述三者的評估結果以獲得各項顧客需求的重要性評比。此外，本研究延伸Wasserman 有關關係強度的正規化模式，考慮顧客需求的相關強度，發展出顧客需求與設計需求之新的正規化關係強度模式以改善現行模式的合理性。本研究利用二元模糊語意發展新產品設計規劃之品質機能展開建構模式將有助於掌握顧客需求與獲得較合理之關係矩陣的整合評估結果，將有助於加速新產品開發的時程，除提高企業的競爭力外，更能有效滿足顧客對新產品設計規劃的滿意度。本計劃以一個半導體封裝案例說明所提出之模式的實用性。
Quality function deployment (QFD) is an analytic systematically tool in processing new product design planning in order to maximum customer satisfaction. Unlike in conventional practice, where the assessments are measured using numeric value or fuzzy number, this study adopts the 2-tuple linguistic representational approaches to express the assessments in house of quality (HOQ) to overcome loss of information in existing ways. In HOQ construction, this study adopts the Delphi method to evaluate consensus results of three items, include importance, sales point, and worth to determine the importance score to each customer requirements (CR) based on current and potential customers’ responses with 2-tuple linguistic representation approaches. An ordered weighted geometric averaging (OWGA) operator is applied to aggregate the importance score of each CR the the consensus assessments of importance, sales point and worth. In addition, different from prior studies, this paper extends Wasserman’s model to take the correlations among CRs into account to develop a new normalized relationship between CRs and design requirements (DRs) to improve the rationality of the existing model. The proposed model uses 2-tuple fuzzy linguistic pproach for developing QFD constructing model can improve the understanding to CRs and obtain the more reasonable results for the relationship between CRs and DRs for new product design planning. These benefits will expedite the new product design planning developing schedule, enhance the firm’s competition and customer satisfaction. An example of semi-conductor case is used to demonstrate the applicability of the proposed model.