發(fā)布時(shí)間：2021-04-12 04:46

　　現(xiàn)階段,越來越多的用戶選擇在大規(guī)模開放在線課程（Massive Open Online Course,MOOC）平臺(tái)上學(xué)習(xí)課程。與傳統(tǒng)的課堂教學(xué)不同,MOOC允許用戶突破時(shí)間和空間的限制,在任何時(shí)間地點(diǎn)學(xué)習(xí)課程。同時(shí),MOOC也打破了專業(yè)內(nèi)容的限制,用戶可以根據(jù)自己的喜好選擇不同領(lǐng)域的課程,然而,由于MOOC上的課程眾多,對(duì)MOOC平臺(tái)和平臺(tái)用戶而言,MOOC系統(tǒng)如何幫助用戶找到符合其個(gè)人興趣的高質(zhì)量課程成為一個(gè)亟待解決的問題。推薦系統(tǒng)是一種重要的內(nèi)容過濾系統(tǒng),系統(tǒng)可以基于用戶的歷史行為等信息為用戶提供符合其個(gè)人偏好的優(yōu)質(zhì)內(nèi)容。相比搜索引擎等信息獲取方式,推薦系統(tǒng)可以更直接地分析用戶的歷史行為數(shù)據(jù),挖掘用戶的潛在個(gè)人偏好,為用戶提供更加個(gè)性化的內(nèi)容展現(xiàn)形式。作為MOOC平臺(tái)的重要組成模塊,課程推薦系統(tǒng)的好壞直接影響到了MOOC平臺(tái)的使用體驗(yàn),課程推薦系統(tǒng)的推薦質(zhì)量也會(huì)對(duì)用戶的學(xué)習(xí)效果產(chǎn)生直接的影響,是否能將符合用戶偏好的高質(zhì)量課程推薦給用戶成為了衡量MOOC平臺(tái)質(zhì)量的重要指標(biāo)。在MOOC平臺(tái)上,推薦系統(tǒng)必須解決如下幾個(gè)關(guān)鍵性問題:1.在MOOC系統(tǒng)中,存在著大量的不同領(lǐng)域的課程,如何將... 

【文章來源】：華中師范大學(xué)湖北省 211工程院校 教育部直屬院校

【文章頁數(shù)】：63 頁

【學(xué)位級(jí)別】：碩士

【文章目錄】：
Abstract
1 Introduction
    1.1 Research Background
    1.2 Overseas and Domestic Research Status
    1.3 Our Contribution
2 Literature Review on Recommendation Algorithms
    2.1 Related Work of Recommendation Algorithms
        2.1.1 Item-based Collaborative Filtering
        2.1.2 User-based Collaborative Filtering
    2.2 Performance Evaluation Indexes of Recommendation Algorithm
        2.2.1 Prediction Accuracy Rate of Recommendation
        2.2.2 Prediction Coverage Rate of Recommendation
        2.2.3 Diversity of Recommendation
    2.3 Summary of Existing Methods
    2.4 Summary
3 A User-based Collaborative Filtering Algorithm with Improved PCC
    3.1 The Procedure of User-based Collaborative Filtering with Improved PCC
        3.1.1 Overview of User-CF Algorithm with Improved PCC
        3.1.2 Improved Pearson Correlation Coefficient
        3.1.3 Neighbor Set Construction Based on User Similarity
        3.1.4 Rating Prediction Based on Neighbor Set
    3.2 Experimental Results for User-CF with Improved PCC
        3.2.1 Dataset Preparation for Experiment
        3.2.2 Experimental Design
        3.2.3 Experimental Results
    3.3 Summary
4 Course Recommendation Based on Mixed Similarity with Improved PCC
    4.1 Mixed Similarity Calculation with Multipliers of Improved PCC
        4.1.1 Calculation of User Mixed Similarity
        4.1.2 Similarity Optimization with Multipliers of Improved PCC
        4.1.3 Recommendation Rating Correction Module Based on Quality Index
    4.2 Experimental Results for Personalized Recommendation
        4.2.1 Experimental Design
        4.2.2 Experimental Results
    4.3 Summary
5 Summary and Future Work
    5.1 Summary
    5.2 Future Work
References
Acknowledgements
Appendix A Abstract



本文編號(hào)：3132647