臺灣博碩士論文加值系統

商業廣告行為已經在網路上蓬勃發展，Google和Yahoo也提供平台讓廣告商註冊登記，然後將廣告刊登在搜尋結果頁面或是部落格文章頁面，而其中內容文字比對的廣告方式(Contextual Advertising)會偵測頁面的文字內容，進一步分析後提供廣告列表。但是我們發現文字比對廣告方式提供的廣告，大多數不符合部落格文章作者真正想要的需求，部落格文章通常以情緒性文章居多，而且我們觀察發現作者通常會針對事件在文章中抒發情緒，但是卻不一定會出寫出需求，因此我們提出事件驅動情緒性需求廣告模型。
事件驅動情緒性需求廣告模型包含了三個子模型，分別是情緒模型、隱含需求模型和廣告模型，我們分別使用不同的資源文件來訓練。利用大量的部落格文章來訓練情緒模型，透過此模型將一篇部落格文章歸類到正面情緒或是負面情緒；接著隱含需求模型部分，我們到Yahoo知識+搜尋相關事件和情緒的發問文章，再將回覆中的隱含需求人工挑選出來做為訓練文件，每種不同事件和情緒類別組合會產生一個隱含需求列表；最後是廣告模型，我們搜集Yahoo的廣告資源做為我們的廣告資料庫，並且根據廣告文件的重要性和廣告網站的熱門度來決定廣告的分數。最後，我們提出的事件驅動情緒性需求廣告模型再和向量空間模型兩者的廣告分數做線性組合，最後提供分數較高的廣告。
我們提出的事件驅動情緒性需求的部落格廣告方式，對不同事件相同情緒的文章，能各自推薦適合的廣告；對相同事件不同情緒的部落格文章，也是如此，而我們的方法比Google AdSense的廣告準確率多了大約25%，更證明了我們方法的實用性，就算部落格文章沒有提到需求，我們根據事件和情緒類別的搭配找出可能的隱含需求，也能提供最適合的廣告。

Online commercial advertising has been developed in recent years. Advertisers can set contextual ads about their products or services on the platform which is offered by Google or Yahoo. The method of Contextual Advertising is parsing the content of target page and then showing some ads after analysis. However, we found many ads that are unsatisfied to the actual need of bloggers. The majority of blog posts are affective articles. Bloggers usually express themselves in their posts but not always write their needs down. Therefore, we present Event-driven Emotion-Need-based Advertising Model (EENA Model) to improve the accuracy of Contextual Advertising on blog post.
EENA Model consist of Emotion Model, Hidden Need Model and Advertisement Model. First, Emotion Model can be trained by classifying blog posts into positive or negative class. Second, we gathered some questions and answers which are related with event and emotion classes from Yahoo Answers, and we chose hidden needs manually from answer documents as training documents for Hidden Need Model. Each pair of event and emotion class generated a hidden need list. Third, we crawled 81,304 contextual ads from Yahoo and formulated Advertisement Model according to Significance and Popularity of advertisement. Finally, we had ads in rank order according to score by linear combination of EENA Model and Vector Space Model, and our system suggested the top-N ads.
The Event-driven Emotion-Need-based blog advertising that we proposed can deliver suitable ads in terms of event and emotion. The performance of proposed method is better than Google AdSense about 25%. Even if a blog post without needs, we can find hidden needs by event and emotion and suggest some suitable ads.

摘要 III
Abstract V
誌謝 VII
章節目錄 VIII
表目錄 X
圖目錄 XI
第一章 序論 1
1.1 研究背景 1
1.2 問題與動機 1
1.3 研究方法 4
1.4 研究目標 5
1.5 論文架構 6
第二章 相關研究與文獻 7
2.1 部落格 (Blog) 7
2.2 網路廣告(Online Advertising) 7
2.2.1 贊助商搜尋廣告 (Sponsored Search) 8
2.2.2 文字內容廣告(Contextual Advertising) 9
第三章 研究與方法 11
3.1 研究動機與想法 11
3.1.1 研究動機 11
3.1.2 研究想法 11
3.2事件驅動情緒性需求的部落格廣告方式 13
3.2.1系統架構 13
3.2.2 事件驅動情緒性需求廣告模型 14
(Event-driven Emotion-Need-based Advertising Model) 14
3.3 EENA Model細部說明 17
3.3.1情緒模型 (Emotion Model) 17
3.3.2 隱含需求模型 (Hidden Need Model) 18
3.3.3 廣告模型 (Advertisement Model) 20
3.4 Blog post–Ad Matching 24
第四章 實驗和評估 26
4.1 實驗資料和評估方法 26
4.1.1 實驗資料 26
4.1.2 評估方法 31
4.2 情緒模型實驗 32
4.3 隱含需求模型實驗 36
4.4 廣告模型實驗 40
4.5 Blog post – Ad Matching實驗 43
4.5.1 權重參數調整 43
4.5.2 不同系統效能評比 48
第五章 結論和未來研究方向 54
5.1 結論 54
5.2 未來研究方向 54
參考文獻 56

[1] X. Ni, et al., Exploring in the weblog space by detecting informative and affective articles, in Proceedings of the 16th international conference on World Wide Web. 2007, ACM: Banff, Alberta, Canada. p. 281-290.
[2] C.Yang, K.H. Lin, and H.Chen, Emotion Classification Using Web Blog Corpora, in Proceedings of the IEEE/WIC/ACM International Conference on Web Intelligence. 2007, IEEE Computer Society. p. 275-278.
[3] G. Leshed, and J. Kaye, Understanding how bloggers feel: recognizing affect in blog posts, in CHI '06 extended abstracts on Human factors in computing systems. 2006, ACM. p. 1019-1024.
[4] 楊昌樺, 高虹安, 陳信希, 以部落格語料進行情緒趨勢分析,第十九屆自然語言與語音處理研討會論文集. 2007. p. 205-218.
[5] C. Wang, et al. Understanding Consumers Attitude Toward Advertising. in Eighth Americas Conference on Information System. 2002.
[6] V. Murdock, M. Ciaramita, and V. Plachouras, A noisy-channel approach to contextual advertising, in Proceedings of the 1st international workshop on Data mining and audience intelligence for advertising. 2007, ACM: San Jose, California. p. 21-27.
[7] W.T. Yih, J. Goodman, and V.R. Carvalho, Finding advertising keywords on web pages, in Proceedings of the 15th international conference on World Wide Web. 2006, ACM: Edinburgh, Scotland. p. 213-222.
[8] S. Ravi, et al., Automatic generation of bid phrases for online advertising, in Proceedings of the third ACM international conference on Web search and data mining. 2010, ACM: New York, New York, USA. p. 341-350.
[9] C. Yang, K.H. Lin, and H.H. Chen, Building emotion lexicon from weblog corpora, in Proceedings of the 45th Annual Meeting of the ACL on Interactive Poster and Demonstration Sessions. 2007, Association for Computational Linguistics: Prague, Czech Republic. p. 133-136.
[10] J. Yan, et al., How much can behavioral targeting help online advertising?, in Proceedings of the 18th international conference on World wide web. 2009, ACM: Madrid, Spain. p. 261-270.
[11] T.K. Fan and C.H. Chang, Blogger-centric contextual advertising, in Proceeding of the 18th ACM conference on Information and knowledge management. 2009, ACM: Hong Kong, China. p. 1803-1806.
[12] X. Jin, et al., Sensitive webpage classification for content advertising, in Proceedings of the 1st international workshop on Data mining and audience intelligence for advertising. 2007, ACM: San Jose, California. p. 28-33.
[13] T.K. Fan and C.H. Chang, Sentiment-oriented contextual advertising, in Proceedings of the 31th European Conference on IR Research. 2009. p. 202-215.
[14] Y. Zhang, et al., Learning from multi-topic web documents for contextual advertisement, in Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining. 2008, ACM: Las Vegas, Nevada, USA. p. 1051-1059.
[15] A. Ghose and S. Yang, An empirical analysis of sponsored search performance in search engine advertising, in Proceedings of the international conference on Web search and web data mining. 2008, ACM: Palo Alto, California, USA. p. 241-250.
[16] Y. Zhu, et al., Optimizing search engine revenue in sponsored search, in Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval. 2009, ACM: Boston, MA, USA. p. 588-595.
[17] A. Anagnostopoulos, et al., Just-in-time contextual advertising, in Proceedings of the sixteenth ACM conference on Conference on information and knowledge management. 2007, ACM: Lisbon, Portugal. p. 331-340.
[18] A. Stewart, et al., Discovering information diffusion paths from blogosphere for online advertising, in Proceedings of the 1st international workshop on Data mining and audience intelligence for advertising. 2007, ACM: San Jose, California. p. 46-54.
[19] B. Ribeiro-Neto, et al., Impedance coupling in content-targeted advertising, in Proceedings of the 28th annual international ACM SIGIR conference on Research and development in information retrieval. 2005, ACM: Salvador, Brazil. p. 496-503.
[20] A. Broder, et al., A semantic approach to contextual advertising, in Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval. 2007, ACM: Amsterdam, The Netherlands. p. 559-566.
[21] R. Baeza-Yates and B. Ribeiro-Neto, Modern Information Retrieval. 1998: ACM.
[22] L.A. Adamic, et al., Knowledge sharing and yahoo answers: everyone knows something, in Proceeding of the 17th international conference on World Wide Web. 2008, ACM: Beijing, China. p. 665-674.