文化大學機構典藏 CCUR:Item 987654321/27777
English  |  正體中文  |  简体中文  |  Items with full text/Total items : 46962/50828 (92%)
Visitors : 12443210      Online Users : 846
RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version


    Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/27777


    Title: 模糊協同過濾於網路教材推薦之研究
    Fuzzy Collaborative Filtering for E-learning CourseRecommendationFuzzy Collaborative Filtering for E-learning Course Recommendation
    Authors: 蔡淑慧
    Contributors: 資訊管理研究所碩士在職專班
    Keywords: 網路學習
    模糊協同過濾
    E-learning
    Fuzzy Collaborative Filtering
    Date: 2006
    Issue Date: 2014-08-13 13:09:35 (UTC+8)
    Abstract: 最近網站上的網路學習引起廣泛注意。目前很多系統焦點放在從網頁日誌分析學習者行為,且提供課程推薦給學習者。最成功和廣泛使用的方法是協同過濾。協同過濾方法找出和線上使用者有類似嗜好的其他學習者,且推薦他們喜歡的給線上使用者。我們可以用學習者在某課程單元的瀏覽時間測量相似度。
    在本研究我們建議一個整合協同過濾和模糊集合的課程推薦系統。它找出離散瀏覽時間的尖銳邊緣問題。我們建議的模型包括三部分:「預處理」、「教學單元推論」和「教學單元預測」。「預處理」包括資料清除和每個教學單元模糊值的計算。「教學單元推論」以教學單元的資料計算教學單元間的相似度。「教學單元預測」提供線上使用者未學習的學習單元推薦清單。
    我們運用「立即擊中率」和「後續擊中率」評估推薦品質。實驗結果顯示建議的方法比未使用時間資訊的
    傳統協同過濾效果較好。
    E-learning on the Web has attracted much attention recently. Many of current systems focus on analyzing the learners’ behaviors from the web logs and provide course recommendations for them. One of the most successful and widely used approaches is collaborative filtering (CF). CF approach finds other learners that have shown similar tastes to the current learner and recommends what they have liked to that learner. The likeness can be measured by the viewing time that learns spent on a specific course unit.
    In this paper, we propose a course recommender system that combines collaborative filtering and fuzzy set. It is designed to better address the sharp boundary problem of discretizing the viewing time. The proposed model contains three modules: preprocessing, unit inference and unit prediction. Preprocessing module includes data cleaning and fuzzy value computation of each unit. Unit inference computes the similarity between each pair of units based on their unit profiles. Unit prediction provides a recommendation list of unseen units for the current leaner.
    We use the hit-ratio metric and click-soon-ratio metric to evaluate the quality of a recommendation. The experimental results show that the proposed methods can achieve a better performance than the Traditional CF without time information.
    Appears in Collections:[Department of Information Management & Graduate Institute of Information Management] Thesis

    Files in This Item:

    There are no files associated with this item.



    All items in CCUR are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback