社群媒體平台產生之輿論的影響力之大似乎已超越社群媒體平台設計者當初設計內容篩選演算法時之想像所及，批評者認為，社群媒體平台並不是一理想的言論場域，可能會使人放棄心靈的自由，而在最嚴重的情況下可能會使侵蝕⺠主。為協助解決問題，本研究開發一套輿情觀測工具之雛形，藉由視覺化呈現某時間點某社群媒體場域之輿論情緒化程度、輿論相似度、與其產生之熱詞，以觀察社群媒體場域的輿論生態，而呈現的結果大致與實際閱覽場域評論的主觀感受相符。本研究若繼續精進使工具越發完善，能使預防議題討論走向極端化或幫助行銷公關部門即時因應可能爆發的一面倒負評。

The influence of the opinion generated from the social media platforms seems to have exceeded the imagination of the social media platform designers when they were first designing the content filtering algorithm. Critics think that the social media platform is not an ideal sphere, which may make one lose the freedom of mind, and in the worst cases, it can corrode democracy. In order to help solve the problem, this thesis develops a prototype of public opinion observer to observe the public opinion ecology in the social media field by visualizing the sentiment extent of public opinion, the opinion similarity, and the hot terms of a social media sphere at a certain time. The presented result is generally in accordance with the subjective feeling of reading actual comments. Continuing to refine the tooling, this study can prevent issue discussions go to extremes or help the PR department immediately respond to the potential overwhelming negative opinion.

致謝 ii
摘要 iii
Abstract iv
目錄 v
圖次 vii
表次 x
算法次 xi
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究問題 3
1.3 各章簡介 3
第二章 文獻探討 5
2.1 網路公眾與網路輿論 5
2.1.1 公共領域與公眾 5
2.1.2 網路與網路公眾 5
2.2 自然語言處理 9
2.2.1 中文斷詞 9
結巴中文分詞 9
2.2.2 文本分析 10
理性主義與經驗主義 10
情感分析 10
向量陳述 11
詞袋模型 12
Word2Vec 12
Doc2Vec 16
2.3 對社群媒體的自然語言處理應用 18
2.3.1 網路溫度計DailyView 18
2.3.2 媒礦Minedia 20
20 第三章 研究方法 24
3.1 語料蒐集 24
3.1.1 語料蒐集方向 24
3.1.2 語料蒐集內容24
3.1.3 語料蒐集腳本開發 27
3.2 資料前處理 27
3.2.1 語料庫整合、斷詞處理與停止詞過濾 29
3.3 資料分析 31
3.3.1 場域情緒值 32
3.3.2 輿論相似度 32
3.3.3 場域熱詞 34
3.4 資料呈現 35
第四章 分析結果呈現 36
4.1 以同性戀相關議題為例 36
4.1.1 案例ㄧ 36
4.1.2 案例二 39
4.2 以iPhone發售相關新聞為例 40
4.2.1 案例一 41
4.2.2 案例二 43
第五章 研究結論 48
5.1 研究成果總結 48
5.2 研究貢獻與未來展望 48
參考文獻 50

[1] Z. S. Harris, “Distributional structure”, Word, vol. 10, no. 2-3, pp. 146–162, 1954.
[2] H. Arendt, The Human Condition. University of Chicago Press, 1958.
[3] J. R. Searle, “Minds, brains, and programs”, Behavioral and brain sciences, vol. 3, no. 3, pp. 417–424, 1980.
[4] J. A. Fodor, “Representations philosophical essays on the foundations of cognitive science”, 1981.
[5] J. Habermas, The structural transformation of the public sphere: An inquiry into a category of bourgeois society. MIT press, 1991.
[6] M. Maffesoli, The time of the tribes: The decline of individualism in mass society. Sage, 1995, vol. 41.
[7] J. Habermas, Between facts and norms: Contributions to a discourse theory of law and democracy. Mit Press, 1996.
[8] K. Hetherington, Expressions of identity: Space, performance, politics. Sage, 1998.
[9] A. Joinson, “Social desirability, anonymity, and internet-based questionnaires”, Behavior Research Methods, Instruments, & Computers, vol. 31, no. 3, pp. 433–438, 1999.
[10] J. Habermas, “The public sphere: An encyclopedia article”, Media and cultural studies: Keyworks, pp. 102–107, 2001.
[11] W.-Y. Ma and K.-J. Chen, “Introduction to ckip chinese word segmentation system for the first international chinese word segmentation bakeoff”, in Proceedings of the second SIGHAN work- shop on Chinese language processing-Volume 17, Association for Computational Linguistics, 2003, pp. 168–171.
[12] 中央研究院資訊所詞庫小組. (2003). 中文斷詞系統, [Online]. Available: http://ckipsvr.iis.sinica.edu.tw/ (visited on 01/09/2018).
[13] 孫治本, “網路是公共領域或言論叢林?”, 聯合新聞網, Jun. 26, 2004.
[14] L. A. Adamic and N. Glance, “The political blogosphere and the 2004 u.s. election: Divided they blog”, in Proceedings of the 3rd International Workshop on Link Discovery, ser. LinkKDD ’05, Chicago, Illinois: ACM, 2005, pp. 36–43, ISBN: 1-59593-215-1. DOI: 10.1145/1134271.1134277. [Online]. Available: http: //doi.acm.org/10.1145/1134271.1134277.
[15] K. Wallsten, “Political blogs and the bloggers who blog them: Is the political blogosphere and echo chamber”, in American Political Science Association’s Annual Meeting. Washington, DC September, 2005, pp. 1–4.
[16] T. S. N. L. P. Group. (2006). Stanford word segmenter, [Online]. Available: https://nlp.stanford.edu/software/segmenter.shtml (visited on 01/09/2018).
[17] L.-W. Ku and H.-H. Chen, “Mining opinions from the web: Beyond relevance retrieval”, Journal of the American Society for Information Science and Technology, vol. 58, no. 12, pp. 1838– 1850, 2007.
[18] Z. Li and M. Sun, “Punctuation as implicit annotations for chi- nese word segmentation”, Computational Linguistics, vol. 35, no. 4, pp. 505–512, 2009.
[19] 北京清華大學自然語言處理與社會人文計算實驗室. (2009). Thulac:一个高效的中文词法分析工具包, [Online]. Available: http://thulac.thunlp.org/ (visited on 01/09/2018).
[20] A. Pak and P. Paroubek, “Twitter as a corpus for sentiment analysis and opinion mining.”, in LREc, vol. 10, 2010.
[21] R. ehek and P. Sojka, “Software Framework for Topic Modelling with Large Corpora”, English, in Proceedings of the LREC 2010 Workshop on New Challenges for NLP Frameworks, http://is.muni.cz/publication/884893/en, Valletta, Malta: ELRA, May 2010, pp. 45–50.
[22] 林意仁, “由 ptt gossiping 看板看 [網路公眾]”, 文化研究月報, no. 108, pp. 52–70, 2010.
[23] S. Brody and N. Diakopoulos, “Cooooooooooooooollllllllllllll!!!!!!!!!!!!!!: Using word lengthening to detect sentiment in microblogs”, in Proceedings of the conference on empirical methods in natural language processing, Association for Computational Linguistics, 2011, pp. 562–570.
[24] E. Kouloumpis, T. Wilson, and J. D. Moore, “Twitter sentiment analysis: The good the bad and the omg!”, Icwsm, vol. 11, no. 538- 541, p. 164, 2011.
[25] S. J. (fxsjy). (2013). 結巴中文分詞, [Online]. Available: https://github.com/fxsjy/jieba (visited on 01/09/2018).
[26] T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efficient esti- mation of word representations in vector space”, arXiv preprint arXiv:1301.3781, 2013.
[27] C.-C. Chiang, “利用文字探勘技術萃取旅館評價文章之研究 ”, 中山大學資訊管理學系研究所學位論文, pp. 1–52, 2014.
[28] Q. Le and T. Mikolov, “Distributed representations of sentences and documents”, in Proceedings of the 31st International Conference on Machine Learning (ICML-14), 2014, pp. 1188–1196.
[29] X. Rong, “Word2vec parameter learning explained”, arXiv preprint arXiv:1411.2738, 2014.
[30] A. S. Veenstra, M. D. Hossain, and B. A. Lyons, “Partisan media and discussion as enhancers of the belief gap”, Mass Communication and Society, vol. 17, no. 6, pp. 874–897, 2014.
[31] 網. D. View. (2014). 網路溫度計 daily view, [Online]. Available: https://dailyview.tw/ (visited on 01/17/2018).
[32] 李日斌, “探討臺灣網⺠對鄰國的情感 ”, 中山大學資訊管理學系研究所學位論文, pp. 1–66, 2014.
[33] M. Czerny. (2015). Modern methods for sentiment analysis, [Online]. Available: https://districtdatalabs.silvrback.com/modern-methods-for-sentiment-analysis (visited on 01/17/2018).
[34] C.-C. Li. (2016). 媒礦 minedia, [Online]. Available: https://minedia.info/ (visited on 01/17/2018).
[35] A. Summa, B. Resch, and M. Strube, “Microblog emotion classification by computing similarity in text, time, and space”, in Proceedings of the Workshop on Computational Modeling of People’s Opinions, Personality, and Emotions in Social Media (PEOPLES), 2016, pp. 153–162.
[36] S.-M. Wang and L.-W. Ku, “Antusd: A large chinese sentiment dictionary.”, in LREC, 2016.
[37] 夕岸. (Jul. 16, 2016). 夕岸:社群媒體，是否必然走向分眾 極化?, [Online]. Available: https://theinitium.com/article/20160716-opinion-xian-internet/ (visited on 01/09/2018).
[38] R. Booth and A. Hern. (2017). Facebook admits industry could do more to combat online extremism, [Online]. Available: https://www.theguardian.com/technology/2017/sep/20/facebook-admits-industry-could-do-more-to-combat-online-extremism (visited on 01/09/2018).
[39] D. Ingram. (2017). Google and facebook must do more to tackle online extremism and fake news, says world economic forum, [Online]. Available: http://www.independent.co.uk/news/business/news/google-facebook-online-extremism-tackle-fight-white-supremacist-islamist-racism-us-tech-firms-world-a8026806.html (visited on 01/09/2018).
[40] N. Lomas. (2017). Tech giants told to remove extremist content much faster, [Online]. Available: https://techcrunch.com/2017/09/20/tech-giants-told-to-remove-extremist-content-much-faster/ (visited on 01/09/2018).
[41] f22313467 (軍 曹). (2017). [爆 卦] 世 大 運 打 臉 總 整 理, [Online]. Available: https://disp.cc/b/163-adu2 (visited on 01/09/2018).
[42] 馮克芸. (2017). G7 要求網路公司和社群媒體制止網路上極端內 容, [Online]. Available: https://udn.com/news/story/ 6809/2488750 (visited on 01/09/2018).
[43] S. Chakrabarti. (2018). Hard questions: What effect does social media have on democracy?, [Online]. Available: https://newsroom.fb.com/news/2018/01/effect-social-media-democracy/?frame-nonce=bd5e374778 (visited on 02/17/2018).
[44] S. Fiegerman. (2018). Facebook admits social media can ’corrode democracy’, [Online]. Available: http://money.cnn.com/2018/01/22/technology/facebook-democracy-social-media/index.html (visited on 02/17/2018).
[45] R. Price. (2018). George soros calls facebook and google a ’menace’ to society and ’obstacles to innovation’ in blistering attack, [Online]. Available: http://www.businessinsider.com/george-soros-calls-facebook-google-menace-society-obstacles-innovation-2018-1 (visited on 02/17/2018).
[46] O. Solon. (2018). George soros: Facebook and google a menace to society, [Online]. Available: https://www.theguardian.com/business/2018/jan/25/george-soros-facebook-and-google-are-a-menace-to-society (visited on 02/17/2018).
[47] 網. D. View. (2018). 2017 年第三季台灣 youtuber 網紅商品合作 好感排行榜, [Online]. Available: https://dailyview.tw/InsightReport/Detail/20 (visited on 01/17/2018).