互聯網的快速進步，使得社群平台不斷崛起，伴隨而來的是大量的用戶生成內容(User Generated Content, UGC)。UGC包含許多有價值的資訊，本論文專注於挖掘社群平台之電影評論，並透過特徵提取，對評論進行結構化的轉換以及情感分析。本論文使用中研院Ckip Tagger對原始評論資料斷詞並標註詞性，並透過詞頻與點狀互見訊息(Pointwise Mutual Information, PMI)提取特徵詞和意見詞，最終取得6項構面，以建立情感分析模型。以準確度和召回率作為評估情感分析模型好壞的指標，研究結果顯示本論文之情感分析模型準確度在正、負、中向情感類別的判斷上皆有高準確度，平均高達93%，唯獨召回率在各構面的表現參差不齊。根據本論文之研究結果，電影公司可以此作為營銷情報；潛在客戶也可進行消費決策的判斷，達成實務上的貢獻。

The rapid progress of the Internet has led to the continuous rise of social platforms, accompanied by a large amount of User Generated Content (UGC). UGC contains a lot of valuable information. This paper focuses on mining movie reviews on social platforms, and through feature extraction, structural conversion and sentiment analysis of reviews are performed. This paper uses Academia Sinica’s Ckip Tagger to segment the original review data and mark the part of speech, and extract feature words and opinion words through word frequency and Pointwise Mutual Information (PMI), and finally obtain 6 facets to establish a sentiment analysis model . The accuracy and recall rate are used as indicators to evaluate the quality of sentiment analysis models. The research results show that the accuracy of the sentiment analysis model in this paper has high accuracy in judging positive, negative, and moderate emotion categories, with an average of up to 93%. The only exception is the performance of recall rates varies across all facet. According to the research results of this thesis, film companies can use this as marketing intelligence; potential customers can also make judgments on consumption decisions and achieve practical contributions.

目錄
誌謝 I
中文摘要 II
Abstract III
目錄 IV
表目錄 VI
圖目錄 VII
1.緒論 1
1.1研究動機 1
1.2研究目的 2
1.3研究架構 2
2.文獻探討 4
2.1 電子口碑 4
2.1.1 社群平台評論屬性 4
2.2 文字探勘 5
2.2.1 SVM分類原理 5
2.3情感分析 6
2.3.1情感分析方法 7
2.4情感分析應用 10
3.研究方法 15
3.1研究流程 15
3.2資料蒐集 17
3.3資料預處理 19
3.3.1中文斷詞 19
3.3.2資料清理 21
3.4特徵提取 22
3.4.1特徵詞提取 22
3.4.2意見詞提取 23
3.4.3否定詞詞典 25
3.4.4程度副詞詞典 25
3.5情感分析 25
3.6評估模型 26
4.實驗結果 27
4.1實驗資料 27
4.2評論資料集 27
4.3特徵詞提取結果 28
4.4意見詞提取結果 31
4.5情感分析模型與驗證 34
4.5.1情感分析模型 34
4.5.2模型驗證分析 36
5. 結論與建議 44
參考文獻 46
附錄一 50
附錄二 60

[1] Abu Tair, M. M., & El-Halees, A. M. (2012). Mining educational data to improve students' performance: a case study. Mining educational data to improve students' performance: a case study, 2(2).
[2] Alharbi, F. R., & Khan, M. B. (2019). Identifying comparative opinions in Arabic text in social media using machine learning techniques. SN Applied Sciences, 1(3), 213.
[3] Al-Kabi, M. N., Gigieh, A. H., Alsmadi, I. M., Wahsheh, H. A., & Haidar, M. M. (2014). Opinion mining and analysis for Arabic language. IJACSA) International Journal of Advanced Computer Science and Applications, 5(5), 181-195.
[4] Archak, N., Ghose, A., & Ipeirotis, P. G. (2011). Deriving the pricing power of product features by mining consumer reviews. Management science, 57(8), 1485-1509.
[5] Berry, M. W. (2004). Survey of text mining. Computing Reviews, 45(9), 548.
[6] Aggarwal, C. C., & Zhai, C. (Eds.). (2012). Mining text data. Springer Science & Business Media.
[7] Chang, Y. C., Ku, C. H., & Chen, C. H. (2019). Social media analytics: Extracting and visualizing Hilton hotel ratings and reviews from TripAdvisor. International Journal of Information Management, 48, 263-279.
[8] Church, K. W., & Hanks, P. (1990). Word association norms, mutual information, and lexicography. Computational linguistics, 16(1), 22-29.
[9] Do, H. H., Prasad, P. W. C., Maag, A., & Alsadoon, A. (2019). Deep learning for aspect-based sentiment analysis: a comparative review. Expert Systems with Applications, 118, 272-299.
[10] El Defrawi, M., Salah, M., AbdAlAziz, A. M., & Eldin, A. S. (2017). Comparative relation extraction from Arabic opinions. Int J Comput Sci Inf Secur, 15(10).
[11] Gaining competitive intelligence from social media data. Industrial Management & Data Systems.
[12] Ezpeleta, E., Zurutuza, U., & Hidalgo, J. M. G. (2016). Using personality recognition techniques to improve Bayesian spam filtering. Procesamiento del Lenguaje Natural, 57, 125-132.
[13] Hu, M., & Liu, B. (2004, August). Mining and summarizing customer reviews. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 168-177).
[14] Jindal, N., & Liu, B. (2006, July). Mining comparative sentences and relations. In Aaai (Vol. 22, No. 13311336, p. 9).
[15] Jindal, N., & Liu, B. (2008, February). Opinion spam and analysis. In Proceedings of the 2008 international conference on web search and data mining (pp. 219-230).
[16] Liu, B. (2011). Opinion mining and sentiment analysis. In Web Data Mining (pp. 459-526). Springer, Berlin, Heidelberg.
[17] Liu, B. (2012). Sentiment analysis and opinion mining. Synthesis lectures on human language technologies, 5(1), 1-167.
[18] Liu, B., Hu, M., & Cheng, J. (2005, May). Opinion observer: analyzing and comparing opinions on the web. In Proceedings of the 14th international conference on World Wide Web (pp. 342-351).
[19] Li, F., Huang, M., & Zhu, X. (2010, July). Sentiment analysis with global topics and local dependency. In Twenty-Fourth AAAI Conference on Artificial Intelligence.
[20] Madden, A., Ruthven, I., & McMenemy, D. (2013). A classification scheme for content analyses of YouTube video comments. Journal of documentation.
[21] Mudambi, S. M., & Schuff, D. (2010). Research note: What makes a helpful online review? A study of customer reviews on Amazon. com. MIS quarterly, 185-200.
[22] Nam, M., Lee, E., & Shin, J. (2015). A method for user sentiment classification using Instagram hashtags. Journal of Korea Multimedia Society, 18(11), 1391-1399.
computing. IEEE Transactions on Affective Computing, 5(3), 273-291.
[23] Pang, B., Lee, L., & Vaithyanathan, S. (2002, July). Thumbs up?: sentiment classification using machine learning techniques. In Proceedings of the ACL-02 conference on Empirical methods in natural language processing-Volume 10 (pp. 79-86). Association for Computational Linguistics.
[24] Pang, B., & Lee, L. (2008). Opinion mining and sentiment analysis. Foundations and trends in information retrieval, 2(1-2), 1-135.
[25] Pouransari, H., & Ghili, S. (2014). Deep learning for sentiment analysis of movie reviews. Technical report, Stanford University, Tech. Rep.
[26] Prabowo, R., & Thelwall, M. (2009). Sentiment analysis: A combined approach. Journal of Informetrics, 3(2), 143-157.
[27] Struhl, S. (2015). Practical Text Analytics: Interpreting Text and Unstructured Data for Business Intelligence: Kogan Page Publishers.
[28] Tan, S., & Zhang, J. (2008). An empirical study of sentiment analysis for chinese documents. Expert Systems with applications, 34(4), 2622-2629.
[29] Vinciarelli, A., & Mohammadi, G. (2014). A survey of personality
Wang, X., White, L., Chen, X., He, W., Shen, J., Tian, X., ... & Tao, R. (2015).
[30] Wiebe, J., Bruce, R., & O’Hara, T. P. (1999, June). Development and use of a gold-standard data set for subjectivity classifications. In Proceedings of the 37th annual meeting of the Association for Computational Linguistics (pp. 246-253).
[31] Xia, R., Zong, C., & Li, S. (2011). Ensemble of feature sets and classification algorithms for sentiment classification. Information sciences, 181(6), 1138-1152.
[32]Stephen Follows(2018) Genre trends in global film production.Available from : https://stephenfollows.com/
[33] Academia Sinica(2019) Ckip Tagger .
https://github.com/ckiplab/ckiptagger
[34] Academia Sinica(1997)中研院平衡語料庫詞類標記集.
http://asbc.iis.sinica.edu.tw/