臺灣博碩士論文加值系統

現今 Youtube 平台已提供方便的線上學習環境，無論是教育工作者或是學習者皆可以運用此平台教學與學習。然而在教育工作者經營的同時，理解觀眾的想法與問題對於營運的方向有著重要的影響。在過去針對分析線上平台學生情緒傾向之文獻，以非監督式學習建立模型進行主題挖掘。為了提高處理留言資訊的效益，本研究針對 Youtube 教育影片底下之留言，應用 K-means 演算法進行留言主題的分類，並且將模型建立成網頁系統。針對 37 位受測者對於本研究提出之網頁系統進行易用性與使用者互動滿意度調查，並另外邀請專家對於 功能面向進行訪談。結果顯示，受測者對該系統有著良好的分數，而專家認為簡單的操作與相關功能能吸引使用者想使用該服務，而在主題呈現方式上，能夠附註一句總結文字，讓網頁所呈現之資訊更加清晰。

Nowadays, YouTube platform provides a convenient online learning environment where educators and learners can utilize it for teaching and learning purposes. However, understanding the thoughts and questions of the audience has a significant impact on the direction of operations for educators. In the past, literature focused on analyzing the sentiment tendencies of online platform students using unsupervised learning to establish models for topic mining. In order to enhance the efficiency of processing comment information, this study applies the K-means algorithm to classify the topics of comments under YouTube educational videos and develops the model into a web service. A usability and user interaction satisfaction survey was conducted on 37 participants regarding the web service proposed in this study, and experts were invited for interviews regarding the functional aspects. The results showed that the participants gave high scores to the system, and experts believed that the simplicity of operation and relevant functionalities could attract users to utilize the service. In terms of topic presentation, it was suggested to include a summary sentence to make the information presented on the webpage clearer.

第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 2
1.3 研究範圍與限制 2
1.4 研究流程 2
第二章 文獻探討 5
2.1 線上學習之互動 5
2.2 Youtube教育平臺現況 6
2.3 Youtube留言互動 6
2.4 非監督式學習之聚類應用 7
2.5 系統易用性量表與使用者互動滿意度量表 9
2.6 小結 12
第三章 研究方法與系統設計 13
3.1 研究流程 13
3.2 留言分群模型設計 14
3.3 系統設計 20
3.4 實驗設計 32
第四章 實驗結果與討論 41
4.1 分群模型結果與討論 41
4.2 網頁系統案例分析 48
4.3 問卷資料分析 73
4.4 專家訪談結果 93
第五章 結論與未來建議 99
5.1 研究結論 99
5.2 未來研究與建議 100
參考文獻 103
附錄一　實驗知情同意書 107
附錄二　專家訪談大綱 109

王思峰、李昌雄 (2005)。如何增進線上學習成效：線上實務社群的浮現觀點 中山管理評論，13(3)，749-776。 https://doi.org/10.6160/2005.09.08
台灣思科 (2000)。E-Learning 掀起新經濟網路時代教育革命，通訊雜誌，第 81 期，頁 87-93。
永田ゆかり (2021)。資料視覺化設計：設計人最想學的視覺化魔法，將枯燥數據變成好看好懂的圖表。旗標科技股份有限公司。
李妮燁 (2020)。用機器學習分析並預測線上付費課程用戶的學習行為以及高價值用戶其特徵-以Hahow為例。國立臺灣大學。https://hdl.handle.net/11296/fg6k34
施威銘研究室(2020)。tf.Keras 技術者們-必讀!: 深度學習攻略手冊。旗標科技股份有限公司。 https://books.google.com.tw/books?id=GRSdzQEACAAJ
孫彤(2018)。從跨語際到跨文化:YouTube互動評論之跨文化學習。國立清華大學。 https://hdl.handle.net/11296/jqjurc
張鐘尹(1997)。漢語對話中的疑問句。國立臺灣大學。 https://hdl.handle.net/11296/rx8v23
Brightmart (2023)。 为中文自然语言处理领域发展贡献语料。 brightmart. Retrieved May 25 from https://github.com/brightmart/nlp_chinese_corpus
Google (2017)。 台灣影音使用者洞察：掌握 YouTube 潮流的三大新知。 Google. Retrieved May 25 from https://www.thinkwithgoogle.com/intl/zh-tw/marketing-strategies/video/insights-taiwan-video-users/
Ahmad, J. (2016). Technology Assisted Language Learning is a silver bullet for enhancing Language competence and performance: A Case Study. International Journal of Applied Linguistics and English Literature, 5, 118-131. https://doi.org/10.7575/aiac.ijalel.v.5n.7p.118
Bangor, A. K., Phil&Miller, James. (2009). Determining What Individual SUS Scores Mean: Adding an Adjective Rating Scale. J. Usability Stud., 4, 114-123.
Bannan, B. (2002). Computer-Mediated Communication, eLearning, and Interactivity: A Review of the Research. Quarterly Review of Distance Education, 3.
Brooke, J. (1995). SUS: A quick and dirty usability scale. Usability Eval. Ind., 189.
Chenail, R. (2011). YouTube as a Qualitative Research Asset: Reviewing User Generated Videos as Learning Resources. Qualitative Report, 16, 229-235. https://doi.org/10.46743/2160-3715/2011.1048
Chin, J. D., Virginia&Norman, Kent. (1988). Development of an Instrument Measuring User Satisfaction of the Human-Computer Interface. ACM CHIi'.
Dvir, A., Marnerides, A. K., Dubin, R., & Golan, N. (2019, 18-21 Feb. 2019). Clustering the Unknown - The Youtube Case. 2019 International Conference on Computing, Networking and Communications (ICNC), Honolulu, HI, USA, 2019, pp. 402-407. https://doi.org/10.1109/ICCNC.2019.8685364
Gillan C, Papadakos J, Brual J, Harnett N, Hogan A, Milne E, Giuliani ME. (2018). Impact of High-Fidelity E-Learning on Knowledge Acquisition and Satisfaction in Radiation Oncology Trainees. Current Oncology. 25(6):533-538. https://doi.org/10.3747/co.25.4090
Junyi, S. (2020). jieba. Github. https://github.com/fxsjy/jieba
Liu, Z., Wang, T., Pinkwart, N., & Liu, S. K., Lingyun. (2018). An Emotion Oriented Topic Modeling Approach to Discover What Students are Concerned about in Course Forums. https://doi.org/10.1109/ICALT.2018.00047
Microsoft. (2022). Azure Text Analytics client library for Python - version 5.2.1. Microsoft 2023. Retrieved May 25 from https://learn.microsoft.com/en-us/python/api/overview/azure/ai-textanalytics-readme?view=azure-python
Nunnally, J. C. (1978). Psychometric Theory. McGraw-Hill. https://books.google.com.tw/books?id=WE59AAAAMAAJ
Rejito, J., Atthariq, A., & Abdullah, A. S. (2021). Application of text mining employing k-means algorithms for clustering tweets of Tokopedia. Journal of Physics: Conference Series, 1722(1), 012019. https://doi.org/10.1088/1742-6596/1722/1/012019
Sauro, J. (2011). A Practical Guide to the System Usability Scale: Background, Benchmarks & Best Practices. Measuring Usability LLC. https://books.google.com.tw/books?id=BL0kKQEACAAJ
scikit-learn. (2023). 2.3. Clustering. scikit-learn developers (BSD License). Retrieved May 25 from https://scikit-learn.org/stable/modules/clustering.html#k-means
Snelson, C. (2009). Web-Based Video for e-Learning: Tapping into the YouTubeTM Phenomenon. Collective Intelligence and E-Learning 2.0: Implications of Web-Based Communities and Networking, 147-166. https://doi.org/10.4018/978-1-60566-729-4.ch009
Vara, N., Mirzabeigi, M., Sotudeh, H., & Fakhrahmad, S. M. (2022). Application of k-means clustering algorithm to improve effectiveness of the results recommended by journal recommender system. Scientometrics, 127(6), 3237-3252. https://doi.org/10.1007/s11192-022-04397-4
Youtube. (2022). Implementation: Comments. Youtube. Retrieved May 25 from https://developers.google.com/youtube/v3/guides/implementation/comments
Zhang, L., Wang, S., & Liu, B. (2018). Deep learning for sentiment analysis: A survey.Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery,8(4), e1253. https://doi.org/10.1002/widm.1253