目前運用網路(Internet)的人數日與漸增，無論是電腦網路、手機行動網路都與我們的生活息息相關，甚至在人與人的交流、報章雜誌間的資料，都可以使用電腦、手機的網路來進行沒有距離限制的溝通及閱讀。由於網路資訊共享的趨勢，情緒分析成為資料探勘領域的熱門主題，常見的情緒分析多著重於部落格文章、新聞雜誌等比較長篇的文章中，然而近年來微網誌盛行，短文數量也日漸增加，對短文進行分析的探討研究越來越多樣。
因此本文的目的即為建立短文的情緒分析模式，先利用字典法的原理，使用台大NTUSD情緒辭典以及教育部國語辭典建置情緒辭典，蒐集google美食評論後，使用CKIP中文斷詞系統將文章拆解，並與辭典中的字詞比對，計算出文章情緒分數，為了驗證系統的準確性，使用原始評論者所給予的分數、人工評分後的分數，兩者與系統進行比較，實驗中會發現系統與人工評分的分數有著很高的相同度，實驗結果與不同系統的比較也顯示出我們提出的短文情緒分析模型有一定的準確度。

Nowadays, more and more people rely on the internet for information. As the WIFI and cellular play an important role in our life, we can utilize it to achieve various goals regardless the distance, such as communicating with other people or share the latest news.
Due to the trend of information sharing on the internet, Sentiment Analysis become a popular topic of the data mining area. So far, most of Sentiment Analysis researches focus on articles which contain large words and multiple paragraphs. However, the popularity of micro-blog in recent years, how to analysis the short comments is an important work.
Because more users tend to use short comment to share their experiences on the internet, the purpose of this thesis is to build a Sentiment Analysis model for short or extremely short sentence. In this essay, first the dictionary will be developed with NTUSD and dictionary (Ministry of Education made) by using lexicon-based, collect google food reviews and use the CKIP to disassemble the short comment, and with the dictionary word comparison.We analysis the short comment by using the comments from google rating. According to the result, this Sentiment Analysis model could measure short comments more accurately compared with other models.

摘要 i
英文摘要 ii
第一章 前言 - 1 -
第二章 相關研究 - 3 -
第一節 自然語言處理(NLP) - 3 -
第二節 情緒辭典 - 4 -
第三節 中文斷詞系統 - 5 -
第三章 研究方法 - 7 -
第一節 情緒辭典建立 - 8 -
第二節 情緒分數計算 - 13 -
第三節 星等評分 - 16 -
第四章 實驗結果 - 17 -
第一節 實驗資料來源 - 17 -
第二節 實驗評估方式 - 19 -
第三節 原始資料實驗結果 - 20 -
第四節 人工評分與系統比對 - 22 -
第五節 系統比較實驗 - 25 -
第六節 情緒程度驗證 - 27 -
第七節 實驗結論 - 29 -
第五章 結論與未來工作 - 30 -
第六章 參考文獻 - 32 -
附錄1 - 34 -
附錄2 - 35 -
附錄3 - 36 -

[1]中文斷詞系統 http://ckipsvr.iis.sinica.edu.tw/
[2]百度翻譯，https://fanyi.baidu.com/
[3]卓淑玲、陳學志、鄭昭明(2011)，台灣地區華人情緒與相關心理生理資料庫
— 中文情緒詞常模研究，中華心理學刊55卷，4期，493-523
[4]林彩雯、葉介山、郭秉豫(2016),以 Google App 評論為字詞權重調整之情緒
分析系統,第廿七屆國際資訊管理學術研討會，台中市：靜宜大學
[5]孫瑛澤、陳建良、劉峻杰、劉昭麟、蘇豐文(2010)，中文短句之情緒分類，
第廿二屆自然語言與語音處理研討會，南投市：暨南大學
[6]張育蓉.(2012)，用情緒分析於圖書館使用者滿意度評估之研究，圖書資訊學
研究7:2，153-197
[7]教育部國語詞典http://dict.revised.moe.edu.tw/cbdic/index.html
[8]許菱祥(1965)，中文文法，台北市：大中國圖書股份有限公司
[9]陳淑芬、陳力綺(2017)，現代漢語否定詞「不」和「沒」的句法、語意和言
談/語用特點及其教學應用，University System of Taiwan Working Papers in Linguistics，9，189-202
[10]結巴中文分詞系統https://github.com/fxsjy/jieba
[11]Bahdanau, D., Cho, K. & Bengio, Y. (2015)， Neural machine translation by
jointly learning to align and translate.，In Proc. International Conference on Learning Representations http://arxiv.org/abs/1409.0473
[12] Chowdury, G.G. (2003)，Natural language processing， Annual Review of
Information Science and Technology, 37: 51-89.
[13] google翻譯，https://translate.google.com.tw/?hl=zh-TW
[14] Hownet(知網) http://www.keenage.com/html/c_index.html
[15] Hsu C and Sandford BA(2007)，The Delphi Technique: making sense of consensus，
Practical Assessment, Research & Evaluation 2007; 12: 1.
[16] Hu MQ, Liu B(2004)，Mining and summarizing customer reviews， In:Procedings
of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, ACM,2004,pp.168-177.
[17] Hui-Hsin Wu(2011), Sentiment analysis using multi-dictionary and commonsense
knowledgebase(Unpublished Master’s thesis). Available from Airiti Library.
[18] Jenq-Haur Wang and Chi-Ching Lee(2011). Unsupervised Opinion Phrase
Extraction and Rating in Chinese Blog Posts，In: SocialCom/PASSAT, Boston, MA, IEEE (2011) 820–823
[19] June-Jei Kuo(2013), A Study on Library Users’ Satisfaction Evaluation Using Sentimental Analysis, Journal of Library and Information Science Research 7(2), 153–197
[20] Koehn, P., Och, F. J., and Marcu, D. (2003). Statistical phrase-based translation. In Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology - Volume 1, NAACL ’03, pages 48–54, Stroudsburg, PA, USA. Association for Computational Linguistics
[21] Ku, L.-W.,H.-H. Chen(2007).Mining Opinions from the Web: Beyond Relevance Retrieval.Journal of American Society for Information Science and Technology, Special Issue on Mining Web Resources for Enhancing Information Retrieval,58,1838-1950
[22] L. Rabiner and B.-H. Juang(1993). Fundamentals of Speech Recognition. Prentice Hall Signal Processing Series.
[23] Lun-Wei Ku and Hsin-Hsi Chen (2007). Mining Opinions from the Web: Beyond Relevance Retrieval. Journal of American Society for Information Science and Technology, Special Issue on Mining Web Resources for Enhancing Information Retrieval, 58(12), pages 1838-1850.
[24] Manning, C. D., & Schutze, H. (1999). Foundations of statistical natural language processing. Cambridge Massachusetts: MIT Press.
[25] Chowdhury, G. G. (2003). Natural language processing. Annual review of information science and technology, 37(1):51–89.
[26] NLPLAB-NTUSD http://academiasinicanlplab.github.io/
[27] Pang B, Lee L, Vaithyanathan S. Thumbs up? Sentiment classification using machine learning techniques. In: Isabelle P, ed. Proc. of the EMNLP 2002. Morristown: ACL, 2002. 79−86.
[28] Taboada, Maite, Julian Brooke, Milan Tofiloski, Kimberly Voll, and Manfred Stede(2011). Lexicon-based methods for sentiment analysis. Computational Linguistics. 37(2): p. 267-307.
[29]Y. Wu, M. Schuster, Z. Chen, Q. V. Le, M. Norouzi, W. Macherey, M. Krikun, Y. Cao, Q. Gao, K. Macherey(2016), et al. Google’s neural machine translation system: Bridging the gap between human and machine translation. arXiv:1609.08144 https://arxiv.org/abs/1609.08144