在網際網路盛行下，線上學習已經是一個普遍的學習模式。大規模網路公開課程（MOOCs）是線上學習中的熱門議題，其公開且透過網路教授課程的特性，使大量的人數可以同時註冊一門課，如何了解學生的學習狀態並據以改善他們在學習平台上的服務體驗變得重要。本研究利用資料探勘技術分類人腦電波資料，希望藉由偵測使用者於線上學習時的腦波狀態，準確分類是否對於課程不理解，讓授課者和使用者可以據以調整學習內容並增進學習成效。本研究採用商業型腦波儀器，測試32個受試者在觀看線上學習影片時的腦波資料，並以兩種資料集、兩種正規化方式、兩種時間窗口、兩種類別標記方式組合，產生的十六個資料模型，訓練並測試決策樹、支援向量機、類神經網路三種分類器，以分類準確度、精確度和查全率來衡量分類結果。本研究測試的結果，產生了一個比過去精確度更高的腦波分類器來分辨學習的狀態我們認為這個研究提供了一個可以應用在真實情境的良好的資料處理方式，可以輔助使用者和授課者在線上學習的過程當中了解不理解之處並據以改進，以提高學習成效。

E-learning becomes an alternative learning mode since the prevalence of the Internet. Especially, the advance of MOOC (Massive Open Online Course) technology enabled a course to accommodate tens of thousands of online learners. How to improve learners’ online learning experiences on MOOC platforms becomes a crucial task for platform providers. This research adopts EEG technology to detect learners’ learning states while they are watching videos in online e-learning activities, hoping to improve their learning outcomes. In this research, we built a system to capture and tag the mental states while subjects are watching online videos and use different normalization methods and time windows to process the data obtained from EEG devices. Finally, we used different supervised learning algorithms to train and test the classifiers and evaluate the results. The results proved that we provide an efficient data processing way to train classifiers and obtain the high accuracy rate comparing with that of previous researches. We consider this system can facilitate users’ self-awareness of learning states in an efficient way while they are in online e-learning activities, and improve their experiences in MOOC platforms.

Table of Contents
Chapter 1 Introduction 1
1.1 Research Background 1
1.1.1 E-learning and Massive Open Online Courses (MOOCs) 1
1.1.2 Electroencephalography (EEG) and Portable EEG Device 2
1.2 Research Motivations 3
1.3 Research Objectives 4
Chapter 2 Literature Review 5
2.1 Massive Open Online Courses 5
2.1.1 Definition and Features 5
2.1.2 The Benefits and Challenges of MOOCs 6
2.2 Electroencephalography 7
2.2.1 Electroencephalography 7
2.2.2 Prior Researches on EEG Classification 11
2.3 Design Science Approaches 14
Chapter 3 Research Methodology 16
Chapter 4 Experimental Design 18
4.1 Subjects 18
4.2 Equipment and Environmental Settings 18
4.3 Experimental Procedure 19
4.4 Materials 24
Chapter 5 Data Analysis 25
5.1 Data Description 25
5.2 Data Preprocessing 26
5.2.1 Global and Local Normalization Methods 26
5.2.2 6 Seconds and 8 Seconds Time Windows 27
5.2.3 1, 4, and 5 Second State Labeling 28
5.2.4 Over-sampling 30
5.3 Explanation of 16 Models 30
5.4 EEG Data Classification 32
5.5 Evaluation 33
Chapter 6 Experimental Results 34
6.1 Classification Results in 16 Models 34
6.2 Comparing the Results between Category A and B Datasets 38
6.3 Comparing the Results in Different Normalization Methods 40
6.4 Comparing the Results in Different Time Windows 41
6.5 Comparing the Results in Different State Labeling 42
6.6 ANOVA Tests on Different Algorithms 43
6.7 Test on Different Datasets 48
6.8 Summary of Evaluation Results 51
Chapter 7 Discussion 53
7.1 Application Scenarios 53
7.2 Limitations and Future Works 54
Chapter 8 Conclusion 56
References 58
Appendix A 62
Appendix B 78

Table of Tables
Table 2.1. Summary of different bands 8
Table 5.1. Feature set and examples 26
Table 5.2. Category A: Original Dataset 31
Table 5.3. Category B: Adjusted Dataset 31
Table 5.4. Settings in classification tree nodes in Orange 32
Table 5.5. SVM parameters in Orange 32
Table 5.6. Settings in Neural Network nodes in Orange 33
Table 5.7. Confusion Matrix 33
Table 6.1. Data size in two classes in Models A1 to A8 35
Table 6.2. Classification results in Models A1 to A8 36
Table 6.3. Data amount in two classes in Models B1 to B8 37
Table 6.4. Classification results in Models B1 to B8 38
Table 6.5. p-value of t-testing on two different datasets 39
Table 6.6. p-value of t-testing on two different normalization methods 40
Table 6.7. p-value of t-testing on two different time windows 41
Table 6.8. p-value of t-testing on two different state labeling 42
Table 6.9. The statistic distribution of accuracy rate of Model A3 43
Table 6.10. The ANOVA of accuracy rate of Model A3 43
Table 6.11. The statistic distribution of accuracy rate of Model A4 44
Table 6.12. The ANOVA of accuracy rate of Model A4 44
Table 6.13. The statistic distribution of accuracy rate of Model A7 44
Table 6.14. The ANOVA of accuracy rate of Model A7 45
Table 6.15. The statistic distribution of accuracy rate of Model A8 45
Table 6.16. The ANOVA of accuracy rate of Model A8 45
Table 6.17. Data size in two classes in Model A2 48
Table 6.18. Classification accuracy in Models A2 48
Table 6.19. p-value of t-testing on different training methods 49
Table 6.20. Data size in two classes in Model A4 50
Table 6.21. Classification accuracy in Models A4 50
Table 6.22. p-value of t-testing on different training methods 51

Table of Figures
Figure 2.1. Information Systems Research Framework (Hevner et al., 2004) 15
Figure 3.1. The design of EEG classification system 16
Figure 4.1. Experimental Procedure 19
Figure 4.2. The system snapshot of choosing a video 21
Figure 4.3. The system snapshot of instructing subjects to close their eyes 21
Figure 4.4. The system snapshot of watching a video 22
Figure 4.5. The system snapshot of instructing subjects to remove the EEG device 22
Figure 4.6. The system snapshot of asking subjects to answer questions 23
Figure 4.7. Playing raining sounds for one minute 23
Figure 5.1 Segmenting 10 seconds videos into five 6 seconds epochs 28
Figure 5.2 Segmenting 10 seconds videos into three 8 seconds epochs 28
Figure 5.3. 1 second state labeling under 6 second time window 29
Figure 5.4. 4 second state labeling under 6 second time window 29
Figure 5.5. 1 second state labeling under 8 second time window 30
Figure 5.6. 5 second state labeling under 8 second time window 30
Figure 6.1. Classification Tree Graph 47
Figure 7.1. An application Scenario 54

An, K. O., Kim, J. B., Song, W. K., & Lee, I. H. (2010, September). Development of an emergency call system using a brain computer interface (BCI). In Biomedical Robotics and Biomechatronics (BioRob), 2010 3rd IEEE RAS and EMBS International Conference on (pp. 918-923). IEEE.
Anderson, N. (2014). Free online AP courses debut on edX Web site. The Washington Post, 2, 12.
Cahn, B. R., & Polich, J. (2006). Meditation states and traits: EEG, ERP, and neuroimaging studies. Psychological bulletin, 132(2), 180.
Chen, X., Barnett, D. R., & Stephens, C. (2013). Fad or future: The advantages and challenges of massive open online courses (MOOCs). In Research-to Practice Conference in Adult and Higher Education (pp. 20-21).
Cooper, S., & Sahami, M. (2013). Reflections on Stanford's MOOCs. Communications of the ACM, 56(2), 28-30.
Coursera. (2013, October 23). In Wikipedia, the free encyclopedia. Retrieved June 15, 2015, from https://en.wikipedia.org/wiki/Coursera
Crowley, K., Sliney, A., Pitt, I., & Murphy, D. (2010, July). Evaluating a brain-computer interface to categorise human emotional response. In 2010 10th IEEE International Conference on Advanced Learning Technologies (pp. 276-278).
DeBoer, J., Stump, G. S., Seaton, D., & Breslow, L. (2013). Diversity in MOOC students’ backgrounds and behaviors in relationship to performance in 6.002 x. In Proceedings of the Sixth Learning International Networks Consortium Conference.
Galán, F. C., & Beal, C. R. (2012). EEG estimates of engagement and cognitive workload predict math problem solving outcomes. In User Modeling, Adaptation, and Personalization (pp. 51-62). Springer Berlin Heidelberg.
Gevins, A., & Smith, M. E. (2003). Neurophysiological measures of cognitive workload during human-computer interaction. Theoretical Issues in Ergonomics Science, 4(1-2), 113-131.
Hammond, D. C. (2007). What is neurofeedback?. Journal of Neurotherapy, 10(4), 25-36
Harmony, T., Fernández, T., Silva, J., Bernal, J., Díaz-Comas, L., Reyes, A., & Rodríguez, M. (1996). EEG delta activity: an indicator of attention to internal processing during performance of mental tasks. International journal of psychophysiology, 24(1), 161-171.
Heinrich, H., Gevensleben, H., & Strehl, U. (2007). Annotation: Neurofeedback–train your brain to train behaviour. Journal of Child Psychology and Psychiatry, 48(1), 3-16.
Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004). Design science in information systems research. MIS quarterly, 28(1), 75-105.
Khosrowabadi, R., Quek, H. C., Wahab, A., & Ang, K. K. (2010, August). EEG-based emotion recognition using self-organizing map for boundary detection. In Pattern Recognition (ICPR), 2010 20th International Conference on (pp. 4242-4245). IEEE.
Kirmizi-Alsan, E., Bayraktaroglu, Z., Gurvit, H., Keskin, Y. H., Emre, M., & Demiralp, T. (2006). Comparative analysis of event-related potentials during Go/NoGo and CPT: decomposition of electrophysiological markers of response inhibition and sustained attention. Brain research, 1104(1), 114-128.
Klimesch, W. (1999). EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain research reviews, 29(2), 169-195.
Kolukuluri, S. (2013). Massive Open Online Courses (Doctoral dissertation, Indian Institute of Technology, Bombay Mumbai).
Kothari, R. A. V. I., & Dong, M. I. N. G. (2001). Decision trees for classification: A review and some new results. Pattern Recognit, 171, 169-184.
Larsen, E. A. (2011). Classification of EEG Signals in a Brain-Computer Interface System.
Lee, J. C., & Tan, D. S. (2006, October). Using a low-cost electroencephalograph for task classification in HCI research. In Proceedings of the 19th annual ACM symposium on User interface software and technology (pp. 81-90). ACM.
Liyanagunawardena, T. R. (2015). Massive Open Online Courses. Humanities, 4(1), 35-41.
Mak, J. N., Chan, R. H., & Wong, S. W. (2013, November). Evaluation of mental workload in visual-motor task: Spectral analysis of single-channel frontal EEG. In Industrial Electronics Society, IECON 2013-39th Annual Conference of the IEEE (pp. 8426-8430). IEEE.
McAuley, A., Stewart, B., Siemens, G., & Cormier, D. (2010). The MOOC model for digital practice.
Merica, H., Blois, R., & Gaillard, J. M. (1998). Spectral characteristics of sleep EEG in chronic insomnia. European Journal of Neuroscience, 10(5), 1826-1834.
Mostow, J., Chang, K. M., & Nelson, J. (2011, January). Toward exploiting EEG input in a reading tutor. In Artificial Intelligence in Education (pp. 230-237). Springer Berlin Heidelberg.
Murugappan, M., Ramachandran, N., & Sazali, Y. (2010). Classification of human emotion from EEG using discrete wavelet transform. Journal of Biomedical Science and Engineering, 3(04), 390.
Pappano, L. (2012). The Year of the MOOC. The New York Times, 2(12), 2012.
Rangaswamy, M., Porjesz, B., Chorlian, D. B., Wang, K., Jones, K. A., Bauer, L. O., ... & Begleiter, H. (2002). Beta power in the EEG of alcoholics. Biological psychiatry, 52(8), 831-842.
Ranky, G. N., & Adamovich, S. (2010, March). Analysis of a commercial EEG device for the control of a robot arm. In Bioengineering Conference, Proceedings of the 2010 IEEE 36th Annual Northeast (pp. 1-2). IEEE.
Rodriguez, C. O. (2012). MOOCs and the AI-Stanford Like Courses: Two Successful and Distinct Course Formats for Massive Open Online Courses. European Journal of Open, Distance and E-Learning.
Ruiz, J. G., Mintzer, M. J., & Leipzig, R. M. (2006). The impact of e-learning in medical education. Academic medicine, 81(3), 207-212.
Sadigh, D., Seshia, S. A., & Gupta, M. (2012, October). Automating exercise generation: A step towards meeting the MOOC challenge for embedded systems. In Proceedings of the Workshop on Embedded and Cyber-Physical Systems Education (p. 2). ACM.
Shute, V. J., Levy, R., Baker, R., Zapata, D., & Beck, J. (2009, July). Assessment and learning in intelligent educational systems: A peek into the future. In Proceedings of the Artificial Intelligence and Education (AIED’09) Workshop on Intelligent Educational Games (pp. 99-109).
Siemens, G. (2013). Massive open online courses: Innovation in education. Open educational resources: Innovation, research and practice, 5.
Spiegelhalder, K., Regen, W., Feige, B., Holz, J., Piosczyk, H., Baglioni, C., ... & Nissen, C. (2012). Increased EEG sigma and beta power during NREM sleep in primary insomnia. Biological psychology, 91(3), 329-333.
Srinivasan, R. (1999). Methods to improve the spatial resolution of EEG. International Journal of Bioelectromagnetism, 1(1), 102-111.
Stytsenko, K., Jablonskis, E., & Prahm, C. (2011, August). Evaluation of consumer EEG device Emotiv EPOC. In MEi: CogSci Conference 2011, Ljubljana.
Szafir, D., & Mutlu, B. (2013, April). ARTFul: adaptive review technology for flipped learning. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 1001-1010). ACM.
Teplan, M. (2002). Fundamentals of EEG measurement. Measurement science review, 2(2), 1-11.
Thompson, G. (1990). How Can Correspondence-Based Distance Education be Improved?: A Survey of Attitudes of Students Who Are Not Well Disposed toward Correspondence Study. International Journal of E-Learning & Distance Education, 5(1), 53-65.
Tu, C. C. (2014). The Classification of Mental Effort with BCI: The Preliminary Effort on Service Design (Master’s thesis, National Tsing Hua University, Hsinchu)
Väisänen, O. (2008). Multichannel EEG methods to improve the spatial resolution of cortical potential distribution and the signal quality of deep brain sources. Tampereen teknillinen yliopisto. Julkaisu-Tampere University of Technology. Publication; 741.
Valenzi, S., Islam, T., Jurica, P., & Cichocki, A. (2014). Individual Classification of Emotions Using EEG. Journal of Biomedical Science and Engineering, 2014.
Waard, I. (2012). “Benefit and challenges of a MOOC” 2012-03-01. http://moocguide,wikispaces,com/2.+Benefits+and+challenges+of+a+MOOC
Waldrop, M. M. (2014). Massive open online courses, aka MOOCs, transform higher education and science.
Wang, H., Li, Y., Hu, X., Yang, Y., Meng, Z., & Chang, K. M. (2013, June). Using EEG to Improve Massive Open Online Courses Feedback Interaction. In AIED Workshops.
Wolpaw, J. R., & McFarland, D. J. (2004). Control of a two-dimensional movement signal by a noninvasive brain-computer interface in humans. Proceedings of the National Academy of Sciences of the United States of America, 101(51), 17849-17854.
Wulf, J., Blohm, I., Leimeister, J. M., & Brenner, W. (2014). Massive Open Online Courses. Business & Information Systems Engineering, 6(2), 111-114.
Yoon, H., Park, S. W., Lee, Y. K., & Jang, J. H. (2013, October). Emotion recognition of serious game players using a simple brain computer interface. In ICT Convergence (ICTC), 2013 International Conference on (pp. 783-786). IEEE.
Yoshida, K., Hirai, F., & Miyaji, I. (2014). Learning System Using Simple Electroencephalograph Feedback Effect During Memory Work. Procedia Computer Science, 35, 1596-1604.
Zhang, D., Zhao, J. L., Zhou, L., & Nunamaker Jr, J. F. (2004). Can e-learning replace classroom learning?. Communications of the ACM, 47(5), 75-79.