本文旨在模糊環境下，為網路服矛供應商建構一套網路聯盟夥伴的選擇模式。文中是以2000年電信總局登記且目前實際營象之17家網路服務供應商為調查對象，參考文獻整理出聯盟夥伴評選準則，再針對實際參與聯盟決策或與其職務有關之主管人員及專業幕僚共67位進行問卷訪查，利用因素分析將十九項準則萃取出五項構面，依序為「相容性」、「行銷資源」、「政府政策」、「人力資源」、「合作關係」等，進而建構出評選聯盟夥伴之層級結構。文中分別利用層級分析法決定關鍵因子權重，並經由文中所提的模糊多評準演算法則：該演算法則是根據三角形模與語言變數兩個主要的觀念，作為評估準則之「重要性」與方案滿足各準則之「適合度」評估之用。經過層級整合後，利用本文修正之Chang and Chen模糊數排序法求出最適的聯盟對象。最後將本文所提之演算法則，實際應用於數位聯合電信公司，進行決策模式之實證分析，以供網路公司經營者實際決策參考。

　　　　　The main purpose of this paper is to propose an algorithm for internet service providers evaluation in selecting strategic alliance partners from Taiwan application service providers viewpoint under fuzzy environment. The fundamental concepts we have adopted include the factor analysis, eigenvector method, fuzzy Delphi method, fuzzy set theory and multi-criteria decision-marking method. We use factor analysis to condense nineteen alliance partners selection sub-criteria into six criteria to construct the hierarchical structure of evaluation of internet service providers. The fuzzy Delphi method is integrated with the eigenvector method to form a set of pooled weights of the extracted criteria. The concepts of triangular fuzzy number and linguistic variables are used to assess the preference ratings of linguistic variable, 'importance' and 'appropriateness'. Through the hierarchy integration, we obtain the final scores of application service providers preference by appraisers. Then we use a revised Chang and Chen's ranking method to rank the final scores of application service providers for choosing the optimal partner. We find the criteria to judge optimal application service providers strategic alliance partners from internet service providers' standpoints in Taiwan include "compatibility", "marketing resource", "government policy", "human resource", "cooperation". After constructing the hierarchical structure of evaluation, we take SEEDNet Telecom as an empirical demonstration and obtain the optimal strategic alliance partners in application service providers.