近年來MOOC課程，讓教師可以從許多學習向度評估學生以及對學習行為追蹤，然而卻有許多學生在學習課程中面臨認知負擔過重及容易迷失學習方向，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。除此之外，當學生透過影片來學習課程時，就算學生看過影片，也不代表他完全了解影片的內容。

因此，我們透過收集學生觀看影片時的動作資料，搭配上練習題的難易度、鑑別度以及學生的能力值，利用深度學習的方式，去得知學生對該影片內容的熟悉程度以及掌握程度並加以量化。接著我們將熟悉度結合知識地圖並生成每位學生專屬的學習知識地圖，除了可以得到每個知識概念的掌握度，系統可以根據學生不熟悉或未掌握概念去推薦影片以及練習題；學生可以根據知識地圖去得知目前的學習狀況並加以調整自己的學習方式；老師及助教也可以透過觀看每個人的知識地圖得知每位學生的學習狀況，並給予適時的幫助。

除此之外，我們也發現有很多學生有一些特定的行為，比如說作弊。因此我們將利用考試範圍的知識概念分數透過Deep Learning的方式去預測每位學生的考試成績，並將預測成績與真實成績做比較，藉此找出特殊行為的學生們。

Many students face cognitive overload and conceptual disorientation in massive open online courses. In fact, even when a student watches a video, it does not mean that he or she completely understands the content. At the same time, most of the current MOOC platforms have few other methods of assessing students besides exams. Therefore, we used the learning logs of videos, the difficulty of exercises, the discrimination of exercises, and the student’s ability, combined with deep learning, to determine the student’s familiarity of video content. We then determined a knowledge node score, which we integrated with a knowledge map. The map could be personalized to recommend videos and exercises according to the knowledge nodes that had a low score. Finally, students could learn where they needed to improve and adjust their learning accordingly. Moreover, teachers could observe students’ knowledge maps to understand their performance and help them accordingly. Additionally, we also found that there are many students have some particular behavior such as cheating. Consequently, we used the knowledge score to predict the student’s performance in exams and compare the predicted scores with the real scores to find out the students with particular behaviors.

Introduction......1
Background and Related Works......7
System Architecture......14
System Implementation......24
Experiment and Result......52
Conclusion and Future Work......67
Bibliography......69

[1] A. Ng and D. Koller, “Coursera,” Retrieved June 29, 2018, from the World
Wide Web:https://zh-tw.coursera.org, 2012.
[2] M. I. of Technology and H. University, “edx,” Retrieved June 29, 2018, from
the World Wide Web:https://www.edx.org, 2012.
[3] S. Thrun, D. Stavens, and M. Sokolsky, “Udacity,” Retrieved June 29, 2018,
from the World Wide Web:https://www.udacity.com, 2012.
[4] M. Wang, J. Peng, B. Cheng, H. Zhou, and J. Liu, “Knowledge visualization
for self-regulated learning.” Journal of Educational Technology & Society,
vol. 14, no. 3, 2011.
[5] 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.
[6] A. Sunar, S. White, N. Abdullah, and H. Davis, “How learners’interactions
sustain engagement: a mooc case study,” IEEE Transactions on Learning
Technologies, 2016.
[7] C. Tekin, J. Braun, and M. van der Schaar, “etutor: Online learning for personalized
education,” in Acoustics, Speech and Signal Processing (ICASSP),
2015 IEEE International Conference on. IEEE, 2015, pp. 5545–5549.
[8] C. Alario-Hoyos, M. Perez-Sanagustin, C. Delgado-Kloos, M. MunozOrganero
et al., “Delving into participants’profiles and use of social tools
in moocs,” IEEE Transactions on Learning Technologies, vol. 7, no. 3, pp.
260–266, 2014.
[9] S. E. Embretson and S. P. Reise, Item response theory. Psychology Press,
2013.
[10] N. T. University, “Sharecourse,” Retrieved April 16, 2017, from the World
Wide Web:http://www.sharecourse.net/sharecourse/, 2012.
[11] J. C. Nesbit and O. O. Adesope, “Learning with concept and knowledge maps:
A meta-analysis,” Review of educational research, vol. 76, no. 3, pp. 413–448,
2006.
[12] H. Fasihuddin, G. Skinner, and R. Athauda, “Knowledge maps in open learning
environments: An evaluation from learners\,” Journal of Information
Technology and Application in Education, vol. 4, pp. 18–29, 2015.
[13] J. H. Lee and A. Segev, “Knowledge maps for e-learning,” Computers &
Education, vol. 59, no. 2, pp. 353–364, 2012.
[14] N.-F. Huang, I.-H. Hsu, C.-A. Lee, H.-C. Chen, J.-W. Tzeng, and T.-T. Fang,
“The clustering analysis system based on students’motivation and learning behavior,” in IEEE 5th Learning with MOOCS(LWMOOCS V), September 2018
[15] 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.
[16] F. Amadieu, T. Van Gog, F. Paas, A. Tricot, and C. Mariné, “Effects of prior
knowledge and concept-map structure on disorientation, cognitive load, and
learning,” Learning and Instruction, vol. 19, no. 5, pp. 376–386, 2009.
[17] S. M. North, R. Richardson, and M. M. North, “To adapt moocs, or not?
that is no longer the question.” Universal Journal of Educational Research,
vol. 2, no. 1, pp. 69–72, 2014.
[18] Y. Chengjie, “Challenges and changes of mooc to traditional classroom teaching
mode,” Canadian Social Science, vol. 11, no. 1, pp. 135–139, 2015.
[19] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” nature, vol. 521, no.
7553, p. 436, 2015.
[20] J. Han and C. Moraga, “The influence of the sigmoid function parameters
on the speed of backpropagation learning,” in International Workshop on
Artificial Neural Networks. Springer, 1995, pp. 195–201.
[21] K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing
human-level performance on imagenet classification,” in Proceedings of the
IEEE international conference on computer vision, 2015, pp. 1026–1034.
[22] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning representations
by back-propagating errors,” nature, vol. 323, no. 6088, p. 533, 1986.
[23] S. Khan, “Knowledge map from khan academy,” Retrieved June
29, 2018, from the World Wide Web:https://www.khanacademy.org/
exercisedashboard, 2007.
[24] F. Xinzhou, “Knowledge map from junyi academy,” Retrieved June
29, 2018, from the World Wide Web:https://www.junyiacademy.org/
exercisedashboard, 2012.
[25] T.-Y. Yang, C. G. Brinton, C. Joe-Wong, and M. Chiang, “Behavior-based
grade prediction for moocs via time series neural networks,” IEEE Journal of
Selected Topics in Signal Processing, vol. 11, no. 5, pp. 716–728, 2017.
[26] X. Li, L. Xie, and H. Wang, “Grade prediction in moocs,” in Computational
Science and Engineering (CSE) and IEEE Intl Conference on Embedded and
Ubiquitous Computing (EUC) and 15th Intl Symposium on Distributed Computing
and Applications for Business Engineering (DCABES), 2016 IEEE Intl
Conference on. IEEE, 2016, pp. 386–392.
[27] G. van Rossum, “Mongodb,” Retrieved June 29, 2018, from the World Wide
Web:www.mongodb.com, 2009.
[28] C. D. Holley and D. F. Dansereau, Spatial learning strategies: Techniques,
applications, and related issues. Academic Press, 2014.
[29] B. Shneiderman, “Inventing discovery tools: combining information visualization
with data mining,” Information visualization, vol. 1, no. 1, pp. 5–12,
2002.
[30] 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 International Conference on Intelligent Information
Hiding and Multimedia Signal Processing. Springer, 2017, pp.
107–114.
[31] 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.
[32] C. G. Brinton and M. Chiang, “Mooc performance prediction via clickstream
data and social learning networks,” in Computer Communications (INFOCOM),
2015 IEEE Conference on. IEEE, 2015, pp. 2299–2307.
[33] J. Qiu, J. Tang, T. X. Liu, J. Gong, C. Zhang, Q. Zhang, and Y. Xue,
“Modeling and predicting learning behavior in moocs,” in Proceedings of the
ninth ACM international conference on web search and data mining. ACM,
2016, pp. 93–102.
[34] T. Sinha, P. Jermann, N. Li, and P. Dillenbourg, “Your click decides your
fate: Inferring information processing and attrition behavior from mooc video
clickstream interactions,” arXiv preprint arXiv:1407.7131, 2014.
[35] F. Chollet, “Keras,” Retrieved June 29, 2018, from the World Wide
Web:https://keras.io/, 2015.
[36] G. van Rossum, “Python,” Retrieved June 29, 2018, from the World Wide
Web:www.python.org, 1990.
[37] Google Brain Team, “Tensorflow,” Retrieved June 29, 2018, from the World
Wide Web:https://github.com/tensorflow/tensorflow, 2015.
[38] V. Oladokun, A. Adebanjo, and O. Charles-Owaba, “Predicting students’
academic performance using artificial neural network: A case study of an
engineering course,” The Pacific Journal of Science and Technology, vol. 9,
no. 1, pp. 72–79, 2008.
[39] A. S. Lan, A. E. Waters, C. Studer, and R. G. Baraniuk, “Sparse factor
analysis for learning and content analytics,” The Journal of Machine Learning
Research, vol. 15, no. 1, pp. 1959–2008, 2014.
[40] Q. Chen, Y. Chen, D. Liu, C. Shi, Y. Wu, and H. Qu, “Peakvizor: Visual
analytics of peaks in video clickstreams from massive open online courses,”
IEEE Transactions on Visualization & Computer Graphics, no. 10, pp. 2315–
2330, 2016.
[41] 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.
[42] S. M. Kendall, “Rank correlation,” Van Nostrand’s Scientific Encyclopedia,
2005.
[43] A. Ronacher, “Flask,” Retrieved June 29, 2018, from the World Wide
Web:https://github.com/pallets/flask, 2010.
[44] Y. Meier, J. Xu, O. Atan, and M. Van der Schaar, “Predicting grades,” IEEE
Transactions on Signal Processing, vol. 64, no. 4, pp. 959–972, 2016.
[45] Z. Ren, H. Rangwala, and A. Johri, “Predicting performance on mooc assessments
using multi-regression models,” arXiv preprint arXiv:1605.02269,
2016.
[46] National Tsinghua University, “Introduction to computer networks,” Retrieved
June 29, 2018, from the World Wide Web:http://www.sharecourse.
net/sharecourse/course/view/courseInfo/1406, 2018.
[47] ——, “Principles of economics,” Retrieved June 29, 2018, from
the World Wide Web:http://www.sharecourse.net/sharecourse/course/view/
courseInfo/1246, 2018.
[48] ——, “Topics on investment,” Retrieved June 29, 2018, from the World Wide
Web:http://www.sharecourse.net/sharecourse/course/view/courseInfo/1577,
2017.