近年來，由於資訊科技的進步，大規模網路開放式課程（MOOCs）於數位學習的研究領域中逐漸流行與普及。 於此同時，MOOCs也帶給高等教育的許多地機會以及挑戰。使用者只需在MOOCs課程平台上註冊一個帳號，即可透過平台便利地接受高等教育的課程。對於機構或教師而言，開班授課變得更為簡單且較傳統教育能吸引更多的學習者參與課程。然而，機構或教師生產高品質的學習資源需要消耗大量的時間與心力。 為了能夠重複利用這些高品質的學習資源，降低重新創建學習資源的成本，Learning Object (LO)的概念被人們所提出，LO是將學習資源進行模組化，使得使用者可以重複利用學習資源。而儲存LO的內容管理系統被稱為學習物件資源庫（LOR），因此，儲存在學習物件資源庫中的LO應該要很容易被檢索。最常見的方式是通過尋找的學習資源間之關聯性，以增加LO的可搜索性。但同時，它也要求使用者具有相關知識，並通過正確的關鍵字搜索。否則，使用者需要一遍又一遍地重複他們的檢索。且隨著知識的與日俱增，不同的專有名詞被提出，如此又大大增長了使用者利用關鍵字檢索所需資源的困難度。 本研究提出了一種工作群組建構機制。本文所提出的機制使用文字探勘技術來分析LOR上使用者群組的相似性，並藉由使用者群組的相似性比較建構出工作群組的原型，並透過使用者對於LO的偏好進行協同過濾，使得我們可以優化這些工作群組原型。對於LOR的使用者而言，他們可以通過這些與自己相關的工作群組找到他們有興趣的資源、降低重新創建教學資源以及檢索所消耗的時間進而提高生產質量。 ;In recent years, Massive Open Online Courses (MOOCs) get popular in the E-learning research domain with the advance of internet technology. At the same time, MOOCs bring to the higher education massive occasion and challenge. The users can conveniently receive the higher education courses by registering an account on the MOOCs platforms. For institutions or teachers are easier to give a course and attract more participants than traditional education. However, producing high-quality learning materials have to consume lots of time and efforts. To reuse learning materials, lower the cost of recreating the materials. Learning Object (LO) concepts have been proposed to the public. The LO is a modular resource that can be re-used easily by users. The content management system which deposited LO is called Learning Objects Repository (LOR), so the LO which stored in the repository should easily be searched by users. The most common way is to increase the discoverability by finding the relevance of materials, but in the meantime, it requires users to have the relevant knowledge and search via correct keywords. Otherwise, they need to repeat their searches over and over again. However, there are numerous of terms be mentioned with the explosion of knowledge. Users are much harder to discover the materials that they want via correct keywords. This paper proposes a working group construction mechanism for users on LOR. The proposed mechanism applies text mining technique to analyze the similarity of groups to construct prototypes of working groups and find the users′ preference about LO base on collaborative filtering to optimize these prototypes. In the other words, for users on the LOR can quickly discover the materials that they are interested via accessing the working groups which related to themselves and reduce the time consumed about re-creating learning materials, improving production quality.
