透過您的圖書館登入
IP:18.222.193.207
  • 學位論文

半導體產業績效評估模式之建構-以封裝、測試廠商為例

Constructing a Model of Performance Evaluation in Taiwan's Semiconductor Package and Testing Factory

指導教授 : 林榮禾
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


2002年政府宣佈推出兩兆雙星計畫,其中即選擇半導體產業為預期產值破新臺幣兆元的關鍵零組件產業,此外,台灣IC封測產業產值目前為世界第一,更大幅領先於世界各國,隨著大陸半導體產業市場的崛起,台灣面臨極大的威脅和挑戰,如何保持目前的優勢和持續領先,實為目前最大的課題,因此企業必須改變績效評估的模式,以適應新的挑戰和競爭環境。 本研究之目的希望建立出更客觀的績效評估模式,藉此提供該產業實行績效評估政策的量化指標及管理參考依據。研究流程如下: (1)歸納整理半導體相關產業績效評估之文獻,找尋適合的投入項和產出項; (2)應用層級分析法(Analytic Hierarchy Process, AHP)和模糊理論(Fuzzy)求算出各投入項和產出項的權重值,以解決權重不客觀的問題; (3)利用資料包絡分析法(Data Envelopment Analysis, DEA)求算各廠商績效並結合保證區域(Assurance Region, AR)模式和模糊權重排序出封測廠在2010年的績效排名並且利用差額變數分析提出改善方針; (4)長期分析方面則利用視窗分析法(Window Analysis)、麥氏生產力指數(Malmquist Productivity Index, MPI) 進行封測廠2008至2010年度的績效評估;(5)利用超級差額變數分析(Super SBM)模式解決多個決策單位之效率值同為1的問題。最後研究發現,透過DEA/AR、模糊權重、Super SBM模式分析使各DMU鑑別度優於傳統DEA模式,且透過視窗分析和麥式生產力指數模式分析後發現,大型封測廠長期穩定性方面比其他中小型廠商來的好。

並列摘要


The"Two Trillion and Twin Star Development Program"was promoted since 2002. The prior promotion is the key components in the semiconductor industry production and its value is expected to exceed NTD one trillion. Although, the revenue of Taiwan’s IC Packaging and Testing industry is currently first in the world, Taiwan faces a great threats and challenges with the rise of the semiconductor industry in China’s market. The most important issue is how to maintain the current advantages and keep the leading. In order to adapt the new challenges and this competitive environment, the companies must change their performance evaluation model. The purpose of this research is to establish a more objective performance evaluation model and provide the reference of quantitative indicators and management to the industry. First, summarized the performance evaluation literature that related to the semiconductor industry and found the suitable input and output items; Second, used Analytic Hierarchy Process and Fuzzy to calculate the weight of input and output items in order to fix the unbalanced weight problems; Third, used the Data Envelopment to calculate the performance of various manufactories and sort the performance rankings of the Packaging and Testing manufactories in 2010 by Assurance Region and Fuzzy weight then used the slack variable analysis to provide the corrective action; Fourth, in the long-tern analysis, used Window Analysis and Malmquist Productivity Index to evaluate the performance evaluation of the Packaging and Testing manufactories from 2008 to 2010; Fifth, used Slacks-Based Measure of Super-Efficiency Model to solve the problem of efficiency value equal to 1 in various manufactories. This research results reveal that DEA/AR、Fuzzy weight and Super SBM models have more discriminatory ability than the basic DEA models. In addition, according to the Window Analysis and Malmquist Productivity Index models, it is found that big IC Packaging and Testing industry manufacturers have more long-term stability than other small and medium manufacturers.

並列關鍵字

Performance evaluation Window Analysis AHP Fuzzy DEA AR MPI Super SBM

參考文獻


4.2009年全球半導體產業現況
6.高強、黃旭男、Toshiyuki Sueyoshi ,管理績效評估-資料包絡分析法,華泰出版社,2003
東吳大學經濟學系,臺北,2006。
15.邱永和、柯慈儀、楊雯如、林藹新,電子業赴大陸投資之績效評估,真理財
19.馮秀鳳,半導體產業經營效率之研究-資料包絡分析法之應用,碩士論文,中原大學會計學系,桃園,2002。

延伸閱讀