以馬可夫鏈探討雛型法決定資訊需求之適應性過程：以決策支援系統爲例

摘要

雛型法軟體發展摸式其目的是要解決傳統生命週期發展資訊系統時，無法應付使用者需求經常變動以及缺乏彈性的缺點。雖然雛型法是一種能確保資訊需求完整與正確的有效方式，但是雛型法的進行並非漫無止盡的，一旦修改雛型系統的成本將會大於其可能之效益。便應結束修改雛型系統的工作。因此系統發展人貝若能事先掌握相關的資訊。對於雛型系統的修改過程便能加以事先控制，以做爲採用此軟體發展方法之參考，而先前文猷並未提出具體解決方案。但根據觀察與分析，雛型法的資訊需求反復確認過程符合馬可夫鏈決策過程。因此，本研究的目的是提出一個以馬可夫鏈決策過程檢視使用雛型法決定资訊需求之適應性過程架構，並以實際開發「封閉型基金投資組合決策支投系統，之雛型實例來驗證此理論架構。最後，期望其結果能對實務界在資訊系統規则或可行性分析階段可能投入時間、成本之估算能有所助益。

Abstract

The purpose of using prototyping is to solve the problems when using traditional SDLC to develop IS for failing to handle the frequent change of user requirements and lack of flexibility. Although prototyping is an effective method for assuring complete and accurate information requirements determination (IRD), the adaptive process of IRD is not being proceeded indefinitely. Once the cost that it consumes is possibly greater than the corresponding benefit, the adaptive process should be terminated. If system analysts are able to know the relevant information of this process in advance, they will gain control of this process and decide whether to adopt the approach or not. The previous researches on this matter were not mentioned. Therefore, the purpose of this study is to present a theoretical framework of applying h1arkov Chain Process to examine the adaptive process of IRD. Subsequently, an example of closed mutual-fund portfolio DSS is developed in order to validate the framework. Finally, the results of this study wish to provide the helpful guidance for information systems planning or feasibility analysis with respect to the estimation of investment in time and cost of practical situations.

關鍵詞:

• 決策支援系統
• 決定資訊辯求
• 雛型法
• 馬可夫鏈。

Key Words:

• DSS
• information requirements determination
• prototyping
• Markov chain