科技政策與計畫之結構評估模式

Abstract

在科技管理領域中，管理者經常在充滿不確定性的決策環境中做決策，尤其在面對如資訊、生物、奈米等新興科技時，決策者常面臨複雜且交錯的決策問題。在解決複雜的問題時，將問題的考量因素及因素間相互的關係予以圖型化，有利於釐清複雜問題中相關的議題及概念。DEMATEL方法主要的目的即在於研究並解決複雜且決策因子交錯的問題。此方法現在已經廣泛被應用於問題分析或工業規劃等領域之中。應用DEMATEL方法時，可以將欲解決的問題予以圖像化，而這圖像亦反應了參與決策的人員對問題的認知情形。但是要得到合適的圖像，門檻值的設定將影響後續對問題的分析或解決方式。在以往，門檻值通常是經由專家們討論之後定案，但在獲得門檻值的共識是一件困難的事，有時候，亦會由研究者自行考量後訂定，而此將導致不同的研究者將得到不同的門檻值。本論文提出最大平均熵差法(Maximum Mean De-Entropy, MMDE)來解決此問題。本論文亦將引用「矽智財交易中心」之政策規劃及「數位學習」之成效評估模式等兩個案例，說明本方法之可行性，並比較本方法與傳統經由專家們決定門檻值間之異同之處。

Most fields in the management of technology experience uncertainty in the environment of decision-making, especially in emerging technology fields. Information technology, biotechnology and nanotechnology are good examples of sectors with complex coordination problems. To deal with complex problems, structuring them through graphical representations and analyzing causal influences can aid in illuminating complex issues, systems, or concepts. The DEMATEL method is a methodology which can be used for researching and solving complicated and intertwined problem groups. The applicability of the DEMATEL method is widespread, ranging from analyzing world problematique decision making to industrial planning. The end product of the DEMATEL process is a visual representation—the impact-relations map—by which respondents organize their own actions in the world. In order to obtain a suitable impact-relations map, an appropriate threshold value is needed to obtain adequate information for further analysis and decision-making. To obtain the reasonable threshold value, with respect to the difficulty of discussions with experts, the researcher may decide upon the value subjectively themselves. This result may differ among researchers. In this dissertation, we propose a method based on the entropy approach, the Maximum Mean De-Entropy (MMDE) algorithm, to achieve this purpose. Using real cases to find the interrelationships between the criteria for planning SIP Mall policy and evaluating effects in E-learning programs as examples, we will compare the results obtained from the respondents and from our method, and discuss that the different impact-relations maps from these two methods.