數位學習（e-Learning）平台係藉由資訊科技做為媒介，由許多異質學習模組組成，像是：測驗評分模組與學習互動模組，提供授課者與學習者更彈性的學習環境。各學習模組的預設功能及目的不同，因此有必要探討e-Learning平台的使用者對於平台中各學習模組評估準則的差異程度。本研究以兩大e-Learning平台為例，「Moodle」及「Wisdom Master」，並以兩平台中共有的活動流程為標的，即「測驗評分」與「學習互動」，找出授課教師對兩模組的評估要素排序，進而產生e-Learning平台評估機制。本研究所提出的評估機制為二階段模型，第一階段中先由文獻中彙整了三項e-Learning平台評估構面，包括：平台技術、平台功能及平台介面。考量各學習模組異質特性，針對「測驗評分」流程則再細分為九項準則及二十項子準則；而針對「學習互動」流程則細分為八項準則及十八項子準則。並於第一階段實地對台灣地區三十位教師進行深度專家訪談，以模糊層級分析法（FAHP）進行分析。進一步的，在第二階段再結合模糊語意於層級分析結果中，將受測者意向經模糊語意轉換後，以關聯法則進行資料分析。分析結果揭露了不同學習流程之評估要素間相對模糊權重、重要性排序及關聯規則的確存在差異。針對「測驗評分」流程，授課教師普遍認為傳輸資料的正確性與連結穩定性兩項準則最為重要；而針對「學習互動」流程，授課教師最為在意的則是數位學習平台是否提供即時交流知識且能夠快速更新主題內容的互動平台。此外，九項關聯規則分析結果亦揭露了在兩組活動流程中，分別與「介面圖文編排」跟「傳輸資料的正確性」相關的衡量準則，亦可供e-Learning平台開發者做為後續e-Learning平台發展與維護之參考依據。

E-Learning platform enables instructors and learners a more flexible learning environment through information technology. Usually, e-Learning platform contains of diverse learning modules, such as online tests module and interactive learning module. The function for each learning module differ from each other, therefore it is necessary to establish the evaluation criteria for e-Learning platform from specific learning module perspective. In this study, we take two learning modules, the online test module and the interactive learning module, of two e-learning platforms (which are major adoption in Taiwan), Moodle and Wisdom Master, as investigation targets. There are two stages to establish the evaluation mechanism for e-Learning platform. In stages one, the evaluation criteria are collected from previous researches and three dimensions are finally aggregated, including: technology, function and interface, for e-Learning platform evaluation. Three dimensions are further divided into 9 factors (contain 20 sub-criteria) for the interactive learning module evaluation and 8 factors (contain 18 sub-criteria) for the online test module evaluation. A deep interview is implemented and all 30 interviewed instructors rank the evaluate criteria for each learning modules. The collected feedback further analysis using the methodology of analytic hierarchy process (AHP).Furthermore, the second stage adopts a fuzzy mechanism (as known as FAHP) to convert the important weight of AHP result in the first stage into discrete scale for association rule generation. The analysis results reveal that the ranking and important weight of evaluation criteria significant differ from learning muddles that supports our view point. For the online test module, two criteria, the correction for networking transmission and the stability for connection, are primary concern. On the other hand, for the interactive learning module, most instructors are focus on the design of interface for knowledge sharing and course material upload. Finally, the association rules also provide some guideline to improve the performance of e-Learning platform from each learning module. 9 association rules reveal the relationships among evaluation criteria particular for the interface design and the connection correction. The system developers can further refer these contributions for e-Learning platform maintenance and performance improvement.