在本篇論文中，我們研究的問題是從部落格中抓取理由並且在社會科學上有所應用。我們把部落格視為一個公開的意見來源地，可以從裡面看到對於事情許多不同的觀點。為了從部落格中找出人對於熱門議題贊成以及反對的理由，我們的系統分成三個主要的部份：第一個部分判別那各是理由那個不是，第二部份判斷理由是贊成還是反對，第三部份則是理由的分群。我們提出了一個非監督式的方法來解決理由的擷取判斷，主要是藉由跟主題高度相關的字作為判斷理由的依據，在擷取出理由以後我們會判斷每個理由的正反，最後再分別針對正反理由來做分群，在實驗的部份我們用了兩個主題來測試我們提出的方法並且有不錯的結果。在論文的最後我們也針對這些部落格文章的作者做了個人資料分析，這對於社會科學有很大的幫助。

In this paper, we study the problem of summarizing reasons from blogsphere for social study. We regard weblogs as a source for collecting non-discrete public opinions, where genuine aspects can be found. To extract the reason inside the blogs, we define three tasks: reason/non-reason classification, polarity identification, and reason summarization. We solve the reason/non-reason classification problem by selecting a set of topic related words and brief the reasons by clustering paragraphs containing aspects after sentiment classification. Initial experiments on two topics show an encouraging result on the proposed framework. In the end of the paper, we also analyze the bloggers’ profiles which can be helpful to social study.

摘要 I
ABSTRACT II
致謝 III
CONTENT IV
FIGURES VI
TABLES VII
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 RELATED WORK 5
2.1 DIFFERENT DOMAIN ON OPINION EXTRACTION AND SUMMARIZATION 5
2.2 PREVIOUS APPROACH ON OPINION EXTRACTION AND SUMMARIZATION 6
2.3 SENTIMENT CLASSIFICATION 7
CHAPTER 3 TASK DEFINITIONS AND ALGORITHMS 9
3.1 PRE-PROCESSING 10
3.2 FILTERING IRRELEVANT BLOGS 10
3.3 REASON/ASPECT EXTRACTION 11
3.4 SENTIMENT CLASSIFICATION 14
3.5 ASPECT CLUSTERING 15
3.6 PROFILE ANALYSIS 16
CHAPTER 4 EXPERIMENTS 17
4.1 BINARY CLASSIFICATION EVALUATION 17
4.2 CLUSTERING EVALUATION 22
4.3 PROFILE ANALYSIS 26
CHAPTER 5 CONCLUSION 28
REFERENCE 29

1.K. Dave, S. Lawrence, and D.M. Pennock. Mining the Peanut Gallery: Opinion Extraction and Semantic Classification of Product Reviews. WWW’03, pp. 519-528.
2.B. Fung, K. Wang and M. Ester. Hierarchical Document Clustering Using Frequent Itemsets. SDM’03, pp. 59-70.
3.M. Hu and B. Liu. Mining and Summarizing Customer Reviews. SIGKDD’04, pp. 168-177.
4.S.-M. Kim and E. Hovy. Determining the Sentiment of Opinions. Coling’04, pp. 1367-1373.
5.L. W. Ku, Y. T. Liang and H. H. Chen. Opinion extraction, summarization and tracking in news and blog corpora. AAAI-2006 Spring Symposium on Computational Approaches to Analyzing Weblogs, pp. 100-107.
6.B. Liu, M. Hu and J. Cheng. Opinion observer: analyzing and comparing opinions on the Web. WWW’05, pp. 342-351.
7.H. Ohno, Y. Kusumura, Y. Hijikata and S. Nishida, Social summarization method for feedback comments in online auction. Syst. Comput. Japan 37, 8 (Jul. 2006), 38-55.
8.B. Pang, L. Lee and S. Vaithyanathan. Thumbs up? Sentiment classification using machine learning techniques. EMNLP’02, pages 79-86.
9.B. Pang and L. Lee. A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. ACL’04, pages 271-278.
10.E. Riloff and J. Wiebe. Learning Extraction Patterns for Subjective Expressions. EMNLP’03, pp. 105-112.
11.H. Takamura, T. Inui and M. Okumura. Extracting Semantic Orientations of Phrases from Dictionary. NAACL-HLT’07, pp. 292–299.
12.P. Turney. Thumbs up or thumbs down? Semantic orientation applied to unsupervised classification of reviews. ACL’02, pp.417-424.
13.H. Yu and V. Hatzivassiloglou. Towards Answering Opinion Questions: Separating Facts from Opinions and Identifying the Polarity of Opinion Sentences. EMNLP’03, pp. 129–136.
14.L. Zhuang, F. Jing and X.Y. Zhu, Movie review mining and summarization. CIKM’06, pp.43-50.
15.Sentence segmentation tool from Cognitive Computation Group in UIUC http://l2r.cs.uiuc.edu/~cogcomp/tools.php
16.http://www.tartarus.org/martin/PorterStemmer/index.html
17.http://www-tsujii.is.s.u-tokyo.ac.jp/GENIA/tagger/
18.http://www.wjh.harvard.edu/~inquirer/
19.http://ddm.cs.sfu.ca/dmsoft/Clustering/fihc_index.html