灰色理論與類神經網路的自組織映射圖網路模型已經很廣泛應用在各種領域的研究上。本研究主要是應用灰關聯分析找出評分資料之間的關係,並據以修補缺陷的網路評分資料,再配合自組織映射圖網路模型將評分資料做自動分類,建構出可保留缺陷評分資料特徵,以輔助完整評分資料分類的評分樣本資料自動分類系統。實驗結果顯示,利用灰關聯分析配合自組織映射圖網路模型分析出來的結果比傳統統計加總結果便於用視覺辨識及輔助決策判斷。本研究成果可以有效協助廠商掌握產品未上市前市場調查使用者的接受度及興趣取向,降低產品開發成本,提高競爭力。
Grey system theory and self-organizing feature map of artificial neural network model have been extensively used in various applications of research. The purpose of applying grey relational analysis and self-organizing feature map in this research is to find the relationship between rating data with grey relational analysis and then to supplement those incomplete data by the proposed method. Furthermore, the self-organizing feature map is used to cluster the given data. It builds a multi-criteria rating clustering system that can save incomplete rating data’s specifications to assist complete rating data in clustering. The results from grey relational analysis and self-organizing feature map provide better visualization effects and are helpful to decision making than basing on traditional statistical analysis. The presented results can be used to help vendors investigate users’ acceptability and preferences before launching a new product, cutting down R&D costs and enhancing the company’s core competency.