隨著社群網路的蓬勃發展，人們己習慣利用網路搜尋引擎等工具尋找他人對特定商品的評價或使用經驗的分享，以作為購物前的參考。許多研究證明了由廣告所引發的情緒在廣告過程中扮演著至關重要的角色，而這也間接使企業主能夠透過某些手段，例如刻意製造與增加正負面評價的文章或匿名評論，來操縱人們對特定品牌的評價。
雖然網路評價是行銷品牌並增加曝光率的一種方法，且這類具有廣告贊助背景的文章內容在網路上已屬常見，但是一般讀者很難僅透過短暫的文字閱讀即可區別出特定文章是否真實且客觀，因此網路使用者時常會被不真實的廣告贊助文章所誤導，此種情況已令使用者對網路上的內容產生嚴重的不信任感，反而連帶影響到大部份部落格文章的公信力。
為了提供更快速客觀的網路文章判讀工具，提供讀者掌握網路文章廣告贊助背景的機率，本論文結合以機器學習為基礎的文字探勘以及維度化的文字情感分析方式試圖從排名前25名的美食部落格文章中找出能夠判別廣告贊助文章的特徵與模式，藉由使用情感維度系統分析文字情感以及訓練情感模型，我們將有機會挖掘出隱藏在廣告贊助文章裡的深層資料並判斷撰文者撰寫該文章的意圖以及真實性。

With the flourishing of social networks, people are accustomed to finding reviews of specific merchandise or sharing of product experience from others as references before purchasing by using tools such as web search engine. Many researches have proved that the emotions aroused by advertisements play a crucial role in the process of advertisement, and it also provided an opportunity for entrepreneurs indirectly to manipulate the evaluation of people toward specific brand through some means, for example, creating and posting positive and negative posts of review or anonymous comments.
Although posting web reviews is a method to market brands and gain publicity, and such articles with sponsorship contexts have become common on the internet, it is hard for common readers to distinguish the actuality and objectivity of specific article through brief reading, so the internet users are often being misled by dishonest sponsored posts. This situation has caused users to have a serious sense of mistrust for contents on the internet which then in turn affected the credibility of most blog posts.
To provide a faster and more objective tool to analyze web articles, assist readers to grasp the probability of sponsorship contexts of web articles, this thesis attempt to find the features and pattern from blog posts of top 25 gourmets by combining machine learning and dimensional text affection analysis method. By using affective dimension system to analyze text affection and train affective model, we will have the opportunity to discover the deep data hidden in sponsored posts and ascertain the author’s intention and actuality of writing these articles.

目錄 III
圖表目錄 V
1. 研究背景與動機 1
1.1. 研究目的 5
1.1.1. 利用文字探勘從文章裡提取具潛在價值的資訊 6
1.1.2. 利用機器學習從文字裡分析和學習規律並對未知資料進行預測 6
1.1.3. 利用情感維度系統分析廣告贊助文章之情感傾向 6
2. 相關研究 8
2.1. 觀眾情緒對廣告成效之影響 8
2.2. 適用於分析文字情感意義的PAD情感維度系統 9
2.3. ANEW 適用於英文單字的情感規範 11
2.4. 將機器學習應用於文字探勘 13
2.4.1. TF-IDF 14
2.4.2. REPTREE 15
2.4.3. 詞袋模型 BAG OF WORDS 17
3. 研究方法 19
3.1. 研究流程 19
3.2. 研究工具 20
3.3. 文章分詞與去除停止詞(STOP WORDS) 24
3.4. 文字情感分析 – 建置PAD情感維度對應系統 25
3.5. 製作情感訓練集 –分析各分詞之情感特徵＆統計文章情感含量 26
3.6. 計算情感分詞佔所有分詞的比例＆訓練並檢驗情感模型 27
3.7. 問卷系統施測說明 28
4. 問卷調查與資料分析 35
4.1. 與機器學習之正確率比較 35
4.2. 人工回應分析 39
5. 結論與未來應用方向 42
5.1. 結論 42
5.2. 未來應用方向 43
參考文獻 44
附錄1-具代表性之廣告贊助文章10篇 47
附錄2-具代表性之一般文章10篇 52

Ashkanasy, N. M., & Cooper, C. L. (2008). Research companion to emotion in organizations: Edward Elgar Cheltenham, UK.
Batra, R., & Ray, M. L. (1986). Affective responses mediating acceptance of advertising. Journal of consumer research, 13(2), 234-249.
Bollen, J., Mao, H., & Pepe, A. (2011). Modeling public mood and emotion: Twitter sentiment and socio-economic phenomena. ICWSM, 11, 450-453.
Bradley, M. M., Greenwald, M. K., Petry, M. C., & Lang, P. J. (1992). Remembering pictures: Pleasure and arousal in memory. Journal of experimental psychology: Learning, Memory, and Cognition, 18(2), 379.
Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: the self-assessment manikin and the semantic differential. Journal of behavior therapy and experimental psychiatry, 25(1), 49-59.
Bradley, M. M., & Lang, P. J. (1999). Affective norms for English words (ANEW): Instruction manual and affective ratings. Retrieved from
Bronner, F. E., Bronner, J. R., & Faasse, J. (2007). In the mood for advertising. International Journal of Advertising, 26(3), 333-355.
Cao, J., Wang, H., Hu, P., & Miao, J. (2008). PAD model based facial expression analysis. Advances in visual computing, 450-459.
Gill, A. J., Gergle, D., French, R. M., & Oberlander, J. (2008). Emotion rating from short blog texts. Paper presented at the Proceedings of the SIGCHI Conference on Human Factors in Computing Systems.
Gupta, V., & Lehal, G. S. (2009). A survey of text mining techniques and applications. Journal of emerging technologies in web intelligence, 1(1), 60-76.
Jayanthi, J., Kaur, G., & Joseph, K. S. (2017). FINANCIAL FORECASTING USING DECISION TREE (REPTree & C4. 5) AND NEURAL NETWORKS (K*) FOR HANDLING THE MISSING VALUES. ICTACT Journal on Soft Computing, 7(3).
Jieba 斷詞工具. Retrieved from https://github.com/fxsjy/jieba
Kemp, E., Bui, M., & Chapa, S. (2012). The role of advertising in consumer emotion management. International Journal of Advertising, 31(2), 339-353.
Kensinger, E. A., & Corkin, S. (2003). Memory enhancement for emotional words: Are emotional words more vividly remembered than neutral words? Memory & cognition, 31(8), 1169-1180.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1998). Emotion, motivation, and anxiety: Brain mechanisms and psychophysiology. Biological psychiatry, 44(12), 1248-1263.
Mehrabian, A. (1980). Basic Dimensions for a General Psychological Theory: Implications for Personality, Social, Environmental, and Developmental Studies: Oelgeschlager, Gunn & Hain.
Mehrabian, A. (1995). Framework for a comprehensive description and measurement of emotional states. Genetic, social, and general psychology monographs.
Mehrabian, A. (1997). Analysis of affiliation-related traits in terms of the PAD temperament model. The Journal of psychology, 131(1), 101-117.
Mehrabian, A., & Blum, J. S. (1997). Physical appearance, attractiveness, and the mediating role of emotions. Current Psychology, 16(1), 20-42.
Mehrabian, A., Wihardja, C., & Ljunggren, E. (1997). Emotional correlates of preferences for situation-activity combinations in everyday life. Genetic, social, and general psychology monographs, 123(4), 461-478.
Midha, N., & Singh, V. (2018). Classification of E-commerce Products Using RepTree and K-means Hybrid Approach. In Big Data Analytics (pp. 265-273): Springer.
Owolabi, A. B. (2009). Effect of consumers mood on advertising effectiveness. Europe's Journal of Psychology, 5(4), 118.
Pham, M. T., Geuens, M., & De Pelsmacker, P. (2013). The influence of ad-evoked feelings on brand evaluations: Empirical generalizations from consumer responses to more than 1000 TV commercials. International Journal of Research in Marketing, 30(4), 383-394.
Poels, K., & Dewitte, S. (2006). How to capture the heart? Reviewing 20 years of emotion measurement in advertising. Journal of Advertising Research, 46(1), 18-37.
Poria, S., Cambria, E., Bajpai, R., & Hussain, A. (2017). A review of affective computing: From unimodal analysis to multimodal fusion. Information Fusion, 37, 98-125.
Ramos, J. (2003). Using tf-idf to determine word relevance in document queries. Paper presented at the Proceedings of the first instructional conference on machine learning.
Ratneshwar, S., Mick, D. G., & Huffman, C. (2003). The Why of Consumption: Contemporary perspectives on consumer motives, goals, and desires (Vol. 1): Psychology Press.
Remington, N. A., Fabrigar, L. R., & Visser, P. S. (2000). Reexamining the circumplex model of affect. Journal of personality and social psychology, 79(2), 286.
Rubin, D. C., & Talarico, J. M. (2009). A comparison of dimensional models of emotion: Evidence from emotions, prototypical events, autobiographical memories, and words. Memory, 17(8), 802-808.
Russell, J. A. (1980). A circumplex model of affect. Journal of personality and social psychology, 39(6), 1161.
Russell, J. A., & Mehrabian, A. (1977). Evidence for a three-factor theory of emotions. Journal of research in Personality, 11(3), 273-294.
Srinivasan, D. B., & Mekala, P. (2014). Mining Social Networking Data for Classification Using REPTree. International Journal of Advance Research in Computer Science and Management Studies, 2(10).
Sriram, B., Fuhry, D., Demir, E., Ferhatosmanoglu, H., & Demirbas, M. (2010). Short text classification in twitter to improve information filtering. Paper presented at the Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval.
Trstenjak, B., Mikac, S., & Donko, D. (2014). KNN with TF-IDF based Framework for Text Categorization. Procedia Engineering, 69, 1356-1364.
Valdez, P., & Mehrabian, A. (1994). Effects of color on emotions. Journal of experimental psychology: General, 123(4), 394.
Yassine, M., & Hajj, H. (2010). A framework for emotion mining from text in online social networks. Paper presented at the Data Mining Workshops (ICDMW), 2010 IEEE International Conference on.
Yu, H., & Hatzivassiloglou, V. (2003). Towards answering opinion questions: Separating facts from opinions and identifying the polarity of opinion sentences. Paper presented at the Proceedings of the 2003 conference on Empirical methods in natural language processing.
Zolotov, V., & Kung, D. (2017). Analysis and Optimization of fastText Linear Text Classifier. arXiv preprint arXiv:1702.05531.
郑璞, 刘聪慧, & 俞国良. (2012). 情绪诱发方法述评. 心理科学进展, 20(1), 45-55.