隨著開放式教育與學習科技持續演進下,大規模開放線上課程(MOOCs)隨之因應而產生。在台灣MOOCs平台也正在蓬勃發展,均一教育平台是目前台灣免費線上教育平台,提供給學生龐大的影片圖書館、互動式挑戰,還讓家長、老師能夠深入了解學生的學習情況。然而,均一平台雖然提供了豐富的教學短片與互動式練習題,但是卻缺乏良好的課程推薦機制,學生可能學習了超出他能力範圍的課程而遇到學習瓶頸,若此時有個機制在學生學習新章節前先行檢測,了解學生有哪些先備知識不足,並且根據不足的部分推薦課程給學生,將可以降低學生遇到瓶頸的狀況。因此本研究提出一個採用模糊邏輯演算法的課程推薦系統來幫助使用者順利地學習新課程,並且藉由TAM2來探討此系統的使用者行為,實驗結果指出研究模型中的八項假說中有五項是支持的,分別為認知有用性與認知易用性會正向影響使用意圖,工作關聯性、認知易用性與產出品質會正向影響認知有用性。另外成果展示性、主觀規範對認知有用性沒有顯著影響,主觀規範對使用意圖也沒有顯著影響,此三項假說不成立。;Massive open online courses (MOOCs) is rising all over the world recently. In Taiwan, there is a free education platform called “Junyi Academy”. It provides learner massive educational videos and interactive exercises, but lack of lesson recommended that may causes learner stuck at bottleneck when starting a new lesson without enough prior knowledge. In this paper we develop a lesson recommender system based on fuzzy logic to help this problem. We use technology acceptance model 2 to probe user′s technology acceptance. The results showed that perceived usefulness, perceived ease of use have a significant impact on intention to use, and job relevance, perceived ease of use, output quality have a significant impact on perceived usefulness.