大規模線上開放式課程(MOOC)被視為是當代教育的破壞式創新，然而隨著MOOC公司的規模近年來逐漸擴大，與日俱增的成本也迫使這些公司去思考永續經營的付費策略。許多公司由免費服務轉為免費增值服務(Freemium)，提供使用者免費的基本服務，若使用者想使用進階服務再額外付費。

本研究旨在研究MOOC使用者的付費意願影響因子，並找出潛在付費用戶，本研究使用兩種資料探勘的方法，分別是邏輯斯迴歸與決策樹。因為東西文化對消費行為的差異，研究受測者主要收集自美國與台灣。我們設計了四種不同的Freemium付費方案作為因變量，分別為時間限制型:一次性付費、時間限制型:月費制、證書限制型:修業證書、證書限制型:成就證書。在變數設計的部分，我們加入了使用者人口資料、MOOC課堂的各項變數、上課動機、MOOC平台的聲譽、MOOC平台的易用性與有用性、師資的聲譽、教材的滿意度與MOOC課程的互動程度作為自變數來找出影響使用者付費意願的關鍵因子。

結果指出「未來學校申請的上課動機」是上述所有Freemium付費方案的顯著正面影響因素。不僅如此，「用戶與其MOOC教師的互動程度」也是影響付費意願的關鍵決定因素。在其他影響因素的方面，「課程的教學內容」、「使用者在過去曾使用過哪些MOOC平台」等影響因子在特定Freemium方案下的付費意願也有顯著影響。例如，雲端計算相關課程是「時間限制型:月費制」與「證書限制型:成就證書」的正面影響因子。然而，若使用者有使用過Cousera的MOOC課程，則他們會傾向不付費給「時間限制型:月費制」與「證書限制型:修業證書」。最後，本研究針對不同的Freemium方案提出建議。

關鍵字：線上開放式課程、資料探勘、決策樹、邏輯斯迴歸、付費意願、證書、數據分析

MOOC, known as massive open online course, are widely seen as a major part of a larger disruptive innovation taking place in higher education. However, since the scale of MOOC became larger and the costs increased, the MOOC providers found it difficult to maintain their business model without charging from their customers. Freemium is a pricing strategy that a basic service or product is provided for free, while users have to pay for additional features or services. MOOC providers attempt to achieve a profitable purpose through adopting freemium strategy.

This research aims to study the key factors impacting MOOC users’ willingness to pay and to identify the potential paid users by two data mining methods, logistic regression and decision tree. Data were collected from both Taiwan and the United States of America because of their distinct culture. We designed four different freemium strategies (time limited: one-off payment, time limited: monthly payment, certificate limited: certificate of completion, certificate limited: certificate of achievement) as the dependent variables to build models. Participants’ demographic profiles, the descriptions of participant’s specific MOOC course, motivation, MOOC platform’s reputation, MOOC platform’s perceived ease-of-use and MOOC platform’s perceived usefulness, faculty’s reputation, material's satisfaction and interaction are analyzed as independent variables to find out the determinants impacting users’ willingness to pay.

The outcomes show that, users’ motivation regarding future university application is significantly a positive influencer of willingness to pay in all of the freemium strategies. Furthermore, users’ interaction with their MOOC faculty is also a critical determinant impacting the willingness to pay. Regarding the other factors that influence the willingness to pay, the study shows that course content and the platform that user had used before the course play important roles in different freemium strategy. For example, for freemium strategy-- monthly payment and freemium strategy-- certificate of achievement, course content mainly about cloud computing was the positive predictor. However, if the user had used Coursera, then they were less willing to pay for the course in freemium strategy-- monthly payment and freemium strategy-- certificate of completion as well.

Keywords: MOOC, freemium, data mining, decision tree, logistic regression, willingness to pay, certificate, data analysis

ABSTRACT V
1 INTRODUCTION 1
2 LITERATURE REVIEW 4
2.1 FREEMIUM AND MOOC 4
2.2 DATA MINING 5
2.3 PRIOR MOOC STUDIES 7
3 METHODOLOGY 9
3.1 DATA 9
3.2 RESEARCH DESIGN 10
3.2.1 Independent variable (IV) 10
3.2.2 Dependent variables (DV) 17
3.3 COMPARISON OF DATA MINING METHODS 21
4 RESULTS 23
4.1 LOGISTIC REGRESSION (LR) 23
4.1.1 Modeling 23
4.1.2 Model interpretation 31
4.2 DECISION TREE 32
4.2.1 Introduction and data description 32
4.2.2 Model interpretation and discussion 33
5 CONCLUSION 47
5.1 BUSINESS IMPLICATION 47
5.1.1 Time limited freemium strategy 48
5.1.2 Certificate limited freemium strategy 51
5.2 LIMITATION AND FUTURE RESEARCH 54
APPENDIX 59

Allen, I. E., & Seaman, C. A. (2007). Likert scales and data analyses. Quality progress, 40(7), 64-65.

Anderson, C. (2009). Free: The future of a radical price. Random House.

Barber, M., Donnelly, K., Rizvi, S., & Summers, L. (2013). An avalanche is coming. Higher Education and the revolution ahead, 73.

Beiman, L., Friedman, J., Olshen, R., & Stone, C. (1984). Classification and regression trees. California: Wadsworth International Group.

Belleflamme, P., & Jacqmin, J. (2016). An economic appraisal of MOOC platforms: business models and impacts on higher education. CESifo Economic Studies, 62(1), 148-169.

Baker, R. M., & Passmore, D. L. (2016). Value and pricing of MOOCs. Education Sciences, 6(2), 14.

Bali, M. (2014). MOOC pedagogy: gleaning good practice from existing MOOCs. Journal of Online Learning and Teaching, 10(1), 44.

Berinsky, A. J., et al. (2011). "Using Mechanical Turk as a subject recruitment tool for experimental research." Submitted for review.

Casler, K., et al. (2013). "Separate but equal? A comparison of participants and data gathered via Amazon’s MTurk, social media, and face-to-face behavioral testing." Computers in Human Behavior 29(6): 2156-2160.

DeBoer, J., Ho, A. D., Stump, G. S., & Breslow, L. (2014). Changing “course” reconceptualizing educational variables for massive open online courses. Educational researcher, 43(2), 74-84.

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS quarterly, 319-340.

de Oña, J., de Oña, R., & Calvo, F. J. (2012). A classification tree approach to identify key factors of transit service quality. Expert Systems with Applications, 39(12), 11164-11171.

Flavián, C., Longás, L., & Lozano, J. (2013). E-learning and market orientation in higher education. Education and Information Technologies, 18(1), 69-83.

Kass, G. V. (1980). An exploratory technique for investigating large quantities of categorical data. Applied statistics, 119-127.

Kim, J. W., Lee, B. H., Shaw, M. J., Chang, H. L., & Nelson, M. (2001). Application of decision-tree induction techniques to personalized advertisements on internet storefronts. International Journal of Electronic Commerce, 5(3), 45-62.

Kotsiantis, S. B., Zaharakis, I., & Pintelas, P. (2007). Supervised machine learning: A review of classification techniques. Emerging artificial intelligence applications in computer engineering, 160, 3-24.

Kung, L. C., & Lee, W. C. (2017). Pricing and Diversification of Massive Online Open Course Platforms.

Loh, W. Y., & Shih, Y. S. (1997). Split selection methods for classification trees. Statistica Sinica, 7(8), 815–840.

Nie, G., Zhang, L., Li, X., & Shi, Y. (2006, December). The Analysis on the Customers Churn of Charge Email Based on Data Mining Take One Internet Company for Example. In Data Mining Workshops, 2006. ICDM Workshops 2006. Sixth IEEE International Conference on (pp. 843-847). IEEE.

Nie, G., Rowe, W., Zhang, L., Tian, Y., & Shi, Y. (2011). Credit card churn forecasting by logistic regression and decision tree. Expert Systems with Applications, 38(12), 15273-15285.

Quinlan, J. R. (1993). C4.5: Programs for machine learning. Los Altos, CA: Morgan Kaufmann.

Shampanier, K., Mazar, N., & Ariely, D. (2007). Zero as a special price: The true value of free products. Marketing science, 26(6), 742-757.

Shapiro, C., & Varian, H. R. (1998). Versioning: the smart way to. Harvard Business Review, 107(6), 107.

Veletsianos, G., Collier, A., & Schneider, E. (2015). Digging deeper into learners' experiences in MOOC s: Participation in social networks outside of MOOC s, notetaking and contexts surrounding content consumption. British Journal of Educational Technology, 46(3), 570-587.

Wu, D. (2009). Supplier selection: A hybrid model using DEA, decision tree and neural network. Expert Systems with Applications, 36(5), 9105-9112.

Walker, S. H., & Duncan, D. B. (1967). Estimation of the probability of an event as a function of several independent variables. Biometrika, 54(1-2), 167-179.

Zhuoxuan, J., Yan, Z., & Xiaoming, L. (2015). Learning behavior analysis and prediction based on MOOC data. Journal of computer research and development, 52(3), 614-28.

Hsieh, P. C. (2017). 探索不同免費增值策略對於付費意願之影響. 清華大學服務科學研究所學位論文, 1-49.

葉國毅. (2015). MOOC 使用者人格特質與其完成率之相關性. 中央大學企業管理學系學位論文, 1-42.