簡易檢索 / 詳目顯示

回結果列表
研究生： 蔡秉翰
Tsai, Pin-Han
論文名稱： 基於深層神經網路之磨課師知識概念評量系統
Concept Assessment System in MOOCs Based on Deep Neural Networks
指導教授： 黃能富
Huang, Nen-Fu
口試委員: 許健平
Sheu, Jang-Ping
陳俊良
Chen, Jiann-Liang
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 通訊工程研究所
Communications Engineering
論文出版年： 2020
畢業學年度： 108
語文別： 英文
論文頁數： 77
中文關鍵詞： 磨課師 、深度學習 、知識地圖 、學習分析 、動態評量
外文關鍵詞： MOOCs, Deep Learning, Knowledge Map, Learning Analysis, Dynamic Assessment
相關次數： 點閱：2  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 近年來MOOC課程的發展，讓教師可以從許多學習向度對學生進行評估，並且追蹤學生的學習行為，然而，卻有許多學生在學習課程中面臨認知負擔過重或是迷失學習方向，也就是學生在學習時，往往會不清楚有哪些東西還沒學、哪些東西即將要學以及目前學的東西是在整體課程中的哪個概念。因此，過去我們實驗室開發了名為"知識地圖"的系統，可以將學習影片與知識概念節點進行對應，就可以讓學生了解自己在學習的過程中，錯失了哪些重要的觀念，以及對相對應的學習影片多加複習。 但是過去這套系統中，只有利用到學生在學習影片上的行為，進行處理與分析，並沒有使用到學生在練習方面的行為。我們認為練習題是檢視學生影片學習成效的一個直接又簡單的指標，因此，在此系統上，新增了關於練習題上的行為處理，並利用深度學習的方式進行分析，得知學生於該練習題的作答行為，是否代表真正地學會相關的知識概念，將其分析結果，顯示在學生個人化的知識地圖上。 除此之外，我們也為此套系統進行了額外的擴增，像是橫跨教育平台的助教系統，只要是合作平台的課程助教，就能夠在該網頁系統上，進行課程申請並提供資訊，就可以獲得分析學生學習行為的服務，包括大量的學習行為圖表，以及前述的知識地圖系統，讓課程助教與教師能夠更方便啟用各種分析服務，並看見分析結果，以及將結果回傳至教學平台上，並給修習該堂課程的學生使用，可以讓學生在 MOOC 教學上獲得更好的回饋。如此一來，此系統的彈性便提高，未來若是要新增服務，便可以很輕易的使用需要的資訊，對於各個教學平台的反饋，也可以更加簡單。

    Many students face cognitive overload and conceptual disorientation on massive open online courses in recent years. In fact, even if a student watches a video, it does not mean that he or she completely understands the content. Therefore, in the past, a system called "Knowledge Map" was developed by our LAB, which can correspond to the learning videos and knowledge concept nodes. So that students can understand what important concepts they have missed during the learning process, and they can review corresponding learning videos more. However, in this system, only the students' behaviors in learning videos were used for processing and analysis, and the students' behaviors in answering exercises were not used. We believe that exercise is a direct and simple indicator for viewing the effectiveness of students' video learning. Therefore, in the proposed system, new behavior processing on exercises and analysis using deep learning methods are added to learn that whether students really understand knowledge concepts by answering the exercises of relevant knowledge concepts. The analysis results are displayed on the student's personalized knowledge map. In addition, we have also made additional expansions to this system. For example, a teaching assistant system that can cross educational platforms. As long as user is a teaching assistant of a cooperative educational platform, user can apply for courses on the web system and provide the information of course. User can acquire services to analyze student learning behavior, including a large number of learning behavior charts, and the aforementioned knowledge map system. So that course teaching assistants and teachers can more easily enable various analysis services, see the analysis results, and return the results to the educational platform for students who take the course. As a result, the flexibility of this system will increase. If we want to add new services in the future, we can easily use the information, and feedback to various teaching platforms quickly.

    Abstract i 中文摘要 iii Acknowledgement iv Contents v List of Figures ix List of Tables xi Chapter 1 Introduction 1 Chapter 2 Background and Related Works 5 2.1 Deep Neural Network . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.2 Learning Method . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 Video Diagnosis with Knowledge Map . . . . . . . . . . . . . . . . 9 2.2.1 Knowledge Map . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2.2 Dynamic Assessment with Artificial Intelligence . . . . . . . 9 2.2.3 Learning Knowledge Map . . . . . . . . . . . . . . . . . . . 10 2.3 Analysis of the Factors Influencing Learners’ Performance Prediction With Learning Analytics . . . . . . . . . . . . . . . . . . . . . 12 2.3.1 Courses Background and Features . . . . . . . . . . . . . . . 13 2.3.2 Analysis Method . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3.3 Experoment and Result . . . . . . . . . . . . . . . . . . . . 17 2.3.3.1 Question 1 . . . . . . . . . . . . . . . . . . . . . . 17 2.3.3.2 Question 2 . . . . . . . . . . . . . . . . . . . . . . 21 2.3.4 Compare this Research with this Thesis . . . . . . . . . . . 24 Chapter 3 System Architecture 26 3.1 MOOCs platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.1.1 Students Activity Record . . . . . . . . . . . . . . . . . . . 28 3.2 AI Diagnosis System of Knowledge Nodes Proficiency . . . . . . . . 30 3.2.1 Exercise Proficiency Diagnosis Module . . . . . . . . . . . . 30 3.2.2 Replace the Score of Learning Adaptability in Video Diagnosis Module . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2.3 Update Proficiency Daily . . . . . . . . . . . . . . . . . . . 31 3.3 Cross-Platform Analysis System . . . . . . . . . . . . . . . . . . . . 33 Chapter 4 System Implementation 34 4.1 Exercise Proficiency Diagnosis . . . . . . . . . . . . . . . . . . . . . 35 4.1.1 Exercise Answer Record Collection . . . . . . . . . . . . . . 35 4.1.2 Exercise Answer Features Calculation . . . . . . . . . . . . 36 4.1.2.1 Exercise Correct Rate in Class . . . . . . . . . . . 37 4.1.2.2 Attempt Counts When Answering Correct for the First Time . . . . . . . . . . . . . . . . . . . . . . 37 4.1.2.3 Watching Related Videos Before/After Answering Correct for the First Time . . . . . . . . . . . . . 38 4.1.2.4 Answering Process Before Getting First Correct Answer . . . . . . . . . . . . . . . . . . . . . . . . 38 4.1.2.5 Counts of Getting Correct Answer . . . . . . . . . 39 4.1.2.6 Students’ Proficiency of Knowledge Concepts . . . 39 4.1.3 Normalization . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.1.4 Correlation Analysis . . . . . . . . . . . . . . . . . . . . . . 40 4.1.5 Construction of DNN Model . . . . . . . . . . . . . . . . . . 42 4.1.5.1 Model Architecture . . . . . . . . . . . . . . . . . 42 4.1.5.2 Proficiency Score Calculation . . . . . . . . . . . . 44 4.2 Exercise Answer Records Feedback to Video Diagnosis Module . . . 45 4.2.1 Method of Adjustment . . . . . . . . . . . . . . . . . . . . . 45 4.3 Cross-Platform Data Analysis Platform . . . . . . . . . . . . . . . . 48 4.3.1 MOOCs Platform Provides Information . . . . . . . . . . . 48 4.3.2 Display the Results on the Data Analysis Platform . . . . . 50 4.3.3 Provide Data Analysis Result on MOOCs Platform . . . . . 54 Chapter 5 Experiment and Result 56 5.1 Accuracy of exercise proficiency prediction . . . . . . . . . . . . . . 56 5.1.1 Courses Information . . . . . . . . . . . . . . . . . . . . . . 58 5.1.2 Experiment: All Courses Data Split in Proportion . . . . . . 59 5.1.2.1 Training Model . . . . . . . . . . . . . . . . . . . . 59 5.1.2.2 Accuracy of the Result . . . . . . . . . . . . . . . 61 5.1.3 Experiment: Choose Parts of Courses for Training Data . . 63 5.1.3.1 Training model . . . . . . . . . . . . . . . . . . . . 63 5.1.3.2 Accuracy of The Result . . . . . . . . . . . . . . . 64 5.2 Result of Exercise Answer Records Feedback to Video Diagnosis . . 66 5.2.1 Learning Adaptability Higher than Default . . . . . . . . . . 66 5.2.2 Learning Adaptability Lower than Default . . . . . . . . . . 67 5.2.3 The Importance of Learning Adaptability . . . . . . . . . . 68 Chapter 6 Conclusion and Future Work 70 6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Bibliography 72

    [1] P. M. Moreno-Marcos, T. Pong, P. J. Muñoz-Merino, and C. Delgado Kloos,
    “Analysis of the factors influencing learners’ performance prediction with
    learning analytics,” IEEE Access, vol. 8, pp. 5264–5282, 2020.
    [2] S. Fauvel, H. Yu, C. Miao, L. Cui, H. Song, L. Zhang, X. Li, and C. Leung, “Artificial intelligence powered moocs: A brief survey,” in 2018 IEEE International Conference on Agents (ICA), 2018, pp. 56–61.
    [3] C. G. Brinton, M. Chiang, S. Jain, H. Lam, Z. Liu, and F. M. F. Wong, “Learning about social learning in moocs: From statistical analysis to generative model,” IEEE Transactions on Learning Technologies, vol. 7, no. 4, pp.346–359, 2014.
    [4] L. Atiaja and R. S. Guerrero-Proenza, “Moocs: Problems and challenges in higher education,” 07 2016.
    [5] S. Fauvel, H. Yu, C. Miao, L. Cui, H. Song, L. Zhang, X. Li, and C. Leung, “Artificial intelligence powered moocs: A brief survey,” in 2018 IEEE International Conference on Agents (ICA), 2018, pp. 56–61.
    [6] N. Huang, C. Chen, J. Tzeng, T. T. Fang, and C. Lee, “Concept assessment system integrated with a knowledge map using deep learning,” in 2018 Learning With MOOCS (LWMOOCS), 2018, pp. 113–116.
    [7] M. Wang, J. Peng, B. Cheng, H. Zhou, and J. Liu, “Knowledge visualization for self-regulated learning,” Educational Technology Society, vol. 14, pp. 28–42, 07 2011.
    [8] J. A. Ruipérez-Valiente, P. J. Muñoz-Merino, C. D. Kloos, K. Niemann, M. Scheffel, and M. Wolpers, “Analyzing the impact of using optional activities in self-regulated learning,” IEEE Transactions on Learning Technologies, vol. 9, no. 3, pp. 231–243, 2016.
    [9] L. Breslow, D. Pritchard, J. DeBoer, G. Stump, A. Ho, and D. Seaton, “Studying learning in the worldwide classroom: Research into edx’s first mooc,” Research in Practice and Assessment, 06 2013.
    [10] T.-J. Haung, “Imitating the brain with neurocomputer a “new”way towards artificial general intelligence,” International Journal of Automation and Computing, pp. 1–12, 05 2017.
    [11] Y. Lecun, Y. Bengio, and G. Hinton, “Deep learning,” Nature Cell Biology, vol. 521, no. 7553, pp. 436–444, May 2015.
    [12] F. Shaheen, B. Verma, and M. Asafuddoula, “Impact of automatic feature extraction in deep learning architecture,” in 2016 International Conference on Digital Image Computing: Techniques and Applications (DICTA), 2016, pp. 1–8.
    [13] J. Han and C. Moraga, “The influence of the sigmoid function parameters on the speed of backpropagation learning,” in From Natural to Artificial Neural Computation, J. Mira and F. Sandoval, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 1995, pp. 195–201.
    [14] K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,” in 2015 IEEE International Conference on Computer Vision (ICCV), 2015, pp. 1026–1034.
    [15] G. Vrbančič, I. Fister jr, and V. Podgorelec, “Parameter setting for deep neural networks using swarm intelligence on phishing websites classification,” International Journal on Artificial Intelligence Tools, vol. 28, p. 1960008, 09 2019.
    [16] H. Yu and B. M. Wilamowski, “Efficient and reliable training of neural networks,” in 2009 2nd Conference on Human System Interactions, 2009, pp. 109–115.
    [17] N.-F. Huang, C.-A. Lee, Y.-W. Huang, P.-W. Ou, H.-H. Hsu, S.-C. Chen, and J.-W. Tzengßer, “On the automatic construction of knowledge-map from handouts for mooc courses,” in Advances in Intelligent Information Hiding and Multimedia Signal Processing, J.-S. Pan, P.-W. Tsai, J. Watada, and L. C. Jain, Eds. Cham: Springer International Publishing, 2018, pp. 107–114.
    [18] “Sharecourse,” Web:http://www.sharecourse.net/sharecourse/, 2012.
    [19] N. T. University, “Nthu mooc platform,” World Wide Web:https://mooc.nthu.edu.tw/, 2019.
    [20] G. van Rossum, “Mongodb,” Retrieved June 29, 2018, from the World Wide Web:www.mongodb.com, 2009.
    [21] Google Developers, “Youtubeapi,” Retrieved July, 2020, from the World Wide Web:https://developers.google.com/youtube/v3/docs, 2015.
    [22] C. Piech, J. Bassen, J. Huang, S. Ganguli, M. Sahami, L. J. Guibas, and J. Sohl-Dickstein, “Deep knowledge tracing,” in Advances in Neural Information Processing Systems, 2015, pp. 505–513.
    [23] q. liu, Z. Huang, Y. Yin, E. Chen, H. Xiong, Y. Su, and G. Hu, “Ekt: Exerciseaware knowledge tracing for student performance prediction,” IEEE Transactions on Knowledge and Data Engineering, pp. 1–1, 2019.
    [24] Y. Luo, Z.-h. Xiao, J.-p. Li, Z. Xie, and X. Xiao, “Learning behavior big data tells us which students can get a mooc course certificate in chinese university mooc,” 01 2019.
    [25] F. Chollet, “Keras,” Retrieved July, 2020, from the World Wide Web:https://keras.io/, 2015.
    [26] G. van Rossum, “Python,” Retrieved July, 2020, from the World Wide Web:www.python.org, 1990.
    [27] Google Brain Team, “Tensorflow,” Retrieved July, 2020, from the World Wide Web:https://github.com/tensorflow/tensorflow, 2015.
    [28] A. Velázquez and S. Assar, “Student learning styles adaptation method based on teaching strategies and electronic media,” Educational Technology Society, vol. 12, pp. 15–29, 10 2009.
    [29] P. J. M. Merino, J. A. R. Valiente, C. Alario-Hoyos, M. Pérez-Sanagustín, and C. D. Kloos, “Precise effectiveness strategy for analyzing the effectiveness of students with educational resources and activities in moocs,” Comput. Hum. Behav., vol. 47, pp. 108–118, 2015.
    [30] V. Chiley, I. Sharapov, A. Kosson, U. Köster, R. Reece, S. S. de la Fuente, V. Subbiah, and M. James, “Online normalization for training
    neural networks,” CoRR, vol. abs/1905.05894, 2019. [Online]. Available: http://arxiv.org/abs/1905.05894
    [31] A. L. Maas, A. Y. Hannun, and A. Y. Ng, “Rectifier nonlinearities improve neural network acoustic models,” in Proc. icml, vol. 30, no. 1, 2013, p. 3.
    [32] S. M. Kendall, “Rank correlation,” Van Nostrand’s Scientific Encyclopedia, 2005.
    [33] National Tsinghua University, “Introduction to iot,” Retrieved July, 2020, from the World Wide Web:https://www.sharecourse.net/sharecourse/course/view/courseInfo/2102, 2020.
    [34] ——, “Principles of economics (i),” Retrieved July, 2020, from the World Wide Web:https://mooc.nthu.edu.tw/cid=10900ECON0001, 2020.
    [35] ——, “Introduction to life science,” Retrieved July, 2020, from the World Wide Web:https://mooc.nthu.edu.tw/cid=10900LIFT0001, 2020.
    [36] ——, “Calculus (i),” Retrieved July, 2020, from the World Wide Web:https://mooc.nthu.edu.tw/cid=10900MATH0001, 2020.
    [37] ——, “General physics (i),” Retrieved July, 2020, from the World Wide Web:https://mooc.nthu.edu.tw/cid=109001PHY0001, 2020.

    QR CODE