管理情境之反應函數的歸納方法

Abstract

近十年來隨著新科技運用與網路社群普及衝擊了傳統的零售商業模式，供應鏈從傳統的批發價格模式逐漸的進入到線上零售模式，供應商議價能力的領導地位也漸漸地轉移到通路商上面，當產業處在這種劇烈變化的情況下具有高度的不確定性，因此供應鏈的管理方式也應該持續演進，通路商在這個競爭環境需要掌握更多的不確定性。通路商是否能經由與供應商簽定兩部收費契約的方式，強化與供應商間的合作關係，在這種通路商佔優勢的供應鏈中發揮更大影響力，影響供應商的決策、提升獲利。如何在這通路商、供應商之間所構成的兩部收費契約中對不同的影響因素產生的結果作系統性歸納並找出適合的策略。管理情境可以將複雜的變數編制納入一個具有一致性、系統性的情境網格空間當中，發展出多種未來可能成形的情境，將可能形成的情境加以分類歸納成數種策略應用。 本研究針對兩階收費契約進行探討，運用決策樹分類後萃取知識來探討其分類依據的方法如何應用於情境分類歸納的議題。研究首先考慮應用案例-通路商在面對多家供應商時的兩部收費契約設計策略，應用資料庫知識發掘的概念建構成情境空間，透過管理情境的概念調整模型中的反應函數制定合適的情境，運用決策樹分類法將不同的情境結果作分類歸納策略。 最終從情境分類結果歸納出兩家供應商間的市占差異即是本研究案例-通路商在面對多家供應商時的兩部收費契約設計策略的關鍵因素。本研究貢獻發現藉著管理情境的概念讓我們在供應鏈契約設計問題中制定合適的情境，決策樹分類工具對結果做出合適的分類，萃取出有用的知識。

The use of new technologies and the popularity of the online community have impacted the traditional retail business model in the past decade. The supply chain gradually moves from the traditional wholesale price model to the online retail model, and the leadership position of the supplier bargaining power is gradually transferred to the distributors. When the industry is highly uncertain in the case of such drastic changes, the management of the supply chain should continue to improve, and the distributors need to grasp more uncertainties in this competitive environment. We discuss whether distributors providers can strengthen their partnerships with suppliers by signing two fee contracts with suppliers and exert greater influence in the channel-dominant supply chain. This paper discuss the application case of Two-part Tariff Contract, and uses the decision tree to extract the knowledge to explore how the classification method can be applied to the topic classification induction. Finally, from the results the difference between the market share of the two suppliers is the key factor of this Two-part Tariff Contract of the distributors in the face of multiple suppliers. The research contribution found that the concept of management context, we can formulate suitable senarios in the supply chain contract design problem, and the decision tree classification tool can appropriately classify the results and extract useful knowledge.