應用遞增式演算法於點對點式數位學習系統之研究

Abstract

在學習過程中，“分享”及“共同參與”對學習者來說，其實是一件很重要的事。近年來數位學習(E-learning)的研究很多，本文介紹一種P2P(peer to peer)式的數位學習系統。它採用P2P分享檔案的觀念及方法，並藉由結合SCORM連結標籤(Metadata)的資訊與學習者存取學習資源的記錄，透過Incremental kNN演算法進行分析，改善現有P2P系統如Napster與Gnutella等系統，對於系統負載、內容特徵歸類(Content Feature Classification)、檔案索引配置(File Index Distribution)及使用者查詢(User Query)的缺點。During the learning process, sharing and participating is the important thing. Efficiently using the learning contents has been an active research topic in developing the educational material. In this paper, we explore an e-learning prototype based on P2P. This e-learning system uses incremental kNN algorithm to analyze the data, combines of the SCORM's metadata information and the records of learners used particular contents. It improves the disadvantages of Napster and Gnutella, and features with content feature classification, file index distribution, overloading, and user's query.