應用演化趨勢預測企業營運策略之研究--以B2C電子商務為例

Abstract

在企業策略創新相關的研究中，系統化創新方法(TRIZ)能夠提供使用者一些思考的方向，讓企業策略創新的過程更為容易有效。然而「演化趨勢」這項工具，並沒有矛盾矩陣這類的輔助工具幫助使用者簡化分析的過程，因此使用者必須逐一檢視每一個演化趨勢，耗時而費力。 本研究提出一倒傳遞類神經網路(Back-Propagation Neural Network)預測模式的建構程序；此模式可建議少數幾個較有可能的演化趨勢供使用者參考。在蒐集並分析數十個B2C類型電子商務的成功創新策略後，所訓練出之類神經網路模式在B2C類型電子商務產業具有不錯的預測能力。

Among the research of Innovation Strategy, only TRIZ can provide users with directions of thinking in order to make the process of Innovation Strategy easier and more effectively. However, Evolution Trends has no sub-tools which can help reduce the analysis procedure as the Contradiction Matrix. For this reason, it takes much time for the users to inspect every step of Evolution Trends. This research proposes a procedure of predictive model to recommend users a few possible Evolution Trends and effectively reduce the number of trends what users have to inspect. After compiling and analyzing several successful innovation strategy of Electronic Commerce, the Back-Propagation Neural Network really have respectable predictive ability in the field of B2C Electronic Commerce.