本研究實作及評估一種自動給定文件類別標題(Automated Document Labeling )的方法。運用外部階層式知識庫，針對自動歸類(Document Clustering)後的文件集，先抽取類別特徵詞，再以上位詞搜尋演算法(Hypernym Search)取得這些詞彙的上位詞做為概念性類別標題詞。舉例而言，假設文件歸類後抽取出特徵詞「桌子」、「椅子」、「床」；則透過上位詞搜尋，將可對應到「家具」，類似這樣的概念性類別標題詞，將更有助於解讀同類文件的主題。
實驗中使用的測試文件集有英文新聞Reuters-21578、國科會的美國專利文件；以及中文新聞FJU-CTC。使用的階層式知識庫有英文的WordNet以及中文的中研院Sinica BOW。
實驗結果，以人工評估此方式產生的標題詞適切比例，並與InfoMap做比較。國科會的美國專利文件其適切比例約為50%，Reuters-21578約為70%，FJU-CTC低於20%。
歸納結果，知識庫的選擇為影響成效最大關鍵。若有適當之知識庫，則本研究採用的方法可獲得穩定而有效的成效。

A cluster labeling algorithm for creating generic titles based on external resources such as WordNet is proposed. Our method first extracts category-specific terms as cluster title words. These titles words are then mapped to generic terms based on a hypernym search algorithm. The proposed method has been evaluated on a patent document collection , a subset of the Reuters-21578 collection, and a subset of the Chinese collection FJU-CTC. Experimental results revealed that our method performs as anticipated. Real-case applications of these generic terms show promising in assisting humans in interpreting the clustered topics. Our method is general enough such that it can be easily extended to use other hierarchical resources for adaptable label generation.

目錄

第一章 緒論…………………………………………………………………………1
第一節 研究背景………………………………………………………………1
第二節 研究動機………………………………………………………………1
第三節 研究目的………………………………………………………………2
第四節 預期研究貢獻…………………………………………………………2
第二章 文獻探討……………………………………………………………………3
第一節 文件自動歸類方法……………………………………………………3
第二節 標題詞選取方法………………………………………………………8
第三節 WordNet……………………………………………………………… 9
第四節 InfoMap………………………………………………………………12
第五節 中央研究院中英雙語知識本體詞網……………………………… 16
第六節 路透社文件集……………………………………………………… 18
第七節 FJU-CTC文件集 …………………………………………………… 20
第八節 美國專利文件……………………………………………………… 22
第三章 研究設計………………………………………………………………… 28
第一節 實驗設計…………………………………………………………… 28
第二節 實驗工具…………………………………………………………… 30
第三節 測試文件集………………………………………………………… 36
第四節 評估方法…………………………………………………………… 37
第五節 研究限制…………………………………………………………… 38
第四章 研究結果與評估………………………………………………………… 40
第一節 專利文件集………………………………………………………… 40
第二節 路透社文件集……………………………………………………… 44
第三節 FJU-CTC文件集 …………………………………………………… 46
第五章 結論與後續研究方向建議 ………………………………………………52
第一節 結論………………………………………………………………… 52
第二節 後續研究…………………………………………………………… 55
參考書目 ……………………………………………………………………………57



圖目錄

圖 1：階層凝聚式歸類……………………………………………………………… 8
圖 2：WORDNET使用者查詢介面示意圖…………………………………………… 11
圖 3：WORDNET查詢結果示意圖 ………………………………………………… 12
圖 4：INFOMAP相關詞彙查詢介面………………………………………………… 14
圖 5：INFOMAP詞彙網路查詢介面………………………………………………… 15
圖 6：INFOMAP詞彙網路查詢結果………………………………………………… 15
圖 7：SINICA BOW詞彙網路查詢介面 …………………………………………… 17
圖 8：SINICA BOW詞彙網路查詢結果 …………………………………………… 18
圖 9：文件自動歸類流程 ………………………………………………………… 29
圖 10：實驗流程圖………………………………………………………………… 29
圖 11：NSC專利文件歸類結果視覺化…………………………………………… 53























表目錄

表 1：REUTERS-21578（YANG版本） 最大與最小的十類文件統計表…………… 20
表 2：FJU-CTC最大與最小的十類文件統計表…………………………………… 22
表 3：三種選詞方法用在三個文件集的評估結果 ……………………………… 41
表 4：專利文件概念性標題詞評估結果………………………………………… 42
表 5：由MI選出的類別特徵詞…………………………………………………… 44
表 6：路透社文件集概念性標題詞評估結果…………………………………… 45
表 7：FJU-CTC文件集標題詞與概念性標題詞評估結果………………………… 47
表 8：NSC專利文件最後歸類結果………………………………………………… 54

一、中文部分

(一)圖書

林傑斌、劉明德、陳湘，「資料採掘與 OLAP 理論與實務」，台北：文魁書局，2002。

(二)期刊

陳光華，"資訊檢索查詢之自然語言處理"，中國圖書館學會會報，第 57 期， 85年 12月，頁 141 - 153 。
曾元顯，"分類不一致對文件自動分類效果的影響"，大學圖書館，9卷1期，2005年3月，頁 11-13
簡立峰，"尋易系統（Csmart）與中文智慧型資訊檢索"，資訊傳播與圖書館學， 3卷 2期， 85年 12月，頁28-37。

二、英文部分

(一)圖書

Wordnet: An Electronic Lexical Database , pp. xviii-xix

(二)期刊

Aggrawal, C. C., & Yu, P. S., "Finding Generalized Projected Clusters in High Dimensional Spaces," Proceedings of the 2000 ACM SIGMOD international conference on Management of data, Dallas, New York: ACM Press, 2000, 70-81.
Cutting, D. R. Karger, J. O. Pedersen, and J. W. Tukey. “Scatter/gather: A cluster-based approach to browsing large document collections,” Proceedings of the 15th ACM-SIGIR Conference, 1992, pp. 318-329.
Dubes, R. C. & Jain, A. K., "Algorithms for Clustering Data," New Jersey: Prentice Hall, 1988.
Frakes, W. B. & Baezay, R., "Information Retrieval: Data Structures and Algorithms," New Jersey: Prentice-Hall, 1992.


Franca Debole and Fabrizio Sebastiani, “An Analysis of the Relative Hardness of Reuters-21578 Subsets” to appear in Journal of the American Society for Information Science and Technology.
Griffith, A., Luckhurst, H. C. & Willet, P., "Using Inter-Document Similarity Information in Document Retrieval Systems," Journal of the American Society for Information Sciences, Vol. 37, No. 1, 1986, pp. 3-11.
Ido Dagan and Ronen Feldman, "Keyword-based browsing and analysis of large document ets," Proceedings of the Symposium on Document Analysis and Information Retrieval (SDAIR-96), Las Vegas, Nevada, 1996.
Joseph B. Kruskal, "Multidimensional Scaling and Other Methods for Discovering Structure," pp. 296-339 in "Statistical Methods for Digital Computers" edited by Kurt Enslein, Anthony Ralston, and Herbert S. Wilf, Wiley: New York, 1977.
Krista Lagus, Samuel Kaski, and Teuvo Kohonen, “Mining Massive Document Collections by the WEBSOM Method,” Information Sciences, Vol 163/1-3, pp. 135-156, 2004.
Lee-Feng Chien, "PAT-Tree Based Keyword Extraction for Chinese Information Retrieval" CM SIGIR 1997.
Liu, T., Liu, S. & Chen, Z., 2003, "An Evaluation on Feature Selection for Text lustering," Proceedings of the Twentieth International Conference on Machine Learning, Washington, CA: AAAI Press, pp. 488-495.
Marti A. Hearst and Jan O. Pedersen, "Reexamining the Cluster Hypothesis: Scatter/
Gather n Retrieval Results," Proceedings of the 19th ACM-SIGIR Conference, 1996, pp. 76-84.
Mehran Sahami, Salim Yusufali, and Michelle Q. W. Baldonaldo, “SONIA: A Service for Organizing Networked Information Autonomously,” Proceedings of the 3rd ACM Conference on Digital Libraries, 1998, pp. 200-209.
Michele Banko, Vibhu O. Mittal, and Michael J. Witbrock, “Headline Generation Based on Statistical Translation,” ACL 2000.
Oren Zamir and Oren Etzioni, “Web document clustering: a feasibility demonstration,” Proceedings of the 21st ACM-SIGIR Conference, 1998, pp. 46-54.
Paul E. Kennedy, Alexander G. Hauptmann, "Automatic title generation for EM," Proceedings of the 5th ACM Conference on Digital Libraries, 2000, pp.
Ron Bekkerman, Ran El-Yaniv, Yoad Winter, Naftali Tishby, “On Feature Distributional Clustering for Text Categorization,” Proceedings of the 24th ACM-SIGIR Conference, 2001, pp.146-153.
Russell Swan and James Allan, "Automatic Generation of Overview Timelines," Proceedings of the 23rd ACM-SIGIR Conference, 2000, pp. 49-56.
Salton, G. & Buckley, C., "Term Weighting Approaches in Automatic Information Retrieval," Journal of Information Proceeding and Management, Vol. 24, No. 5, 1988, pp. 513-524.
Salton, G., "Automatic Text Processing: The Transformation, Analysis, and Retrieval of Information by Computer," New York: Addison-Wesley, 1989.
Steinbach, M., Karypis, G. & Kumar, V., "A Comparison of Document Clustering Techniques," Technical Report 00-034, Computer Science and Engineering, University of Minnesota, 2000.
Wei,C.P.,Hu P.J.&Dong,Y.X.,"Managing Document Categories in E-Commerce Environments:an Evolution-Based Approach,,"European Journal of Information Systems,Vol.11,No.3,2002,pp.208-222
William B. Frakes and Ricardo Baeza-Yates, Information Retrieval: Data Structure and Algorithms, Prentice Hall, 1992.

(三)網路資源

Antti Arppe, "Term Extraction from Unrestricted Text,",1995 <http://www.lingsoft.fi/doc/nptool/term-extraction.html>
David D. Lewis, “Reuters-21578 text categorization test collection, Distribution 1.0” README file (v 1.2), 1997 <http://www.research.att.com/~lewis/>
Document Understanding Conferences<http://www-nlpir.nist.gov/projects/duc/.>
Jean Godby, "Two Techniques for the Identification of Phrases in Full Text," <http://www.oclc.org/oclc/research/publications/review94/part1/twotech.htm>
Jen-Nan Chen, Jyun-Sheng, Chang and Huey-Chyun Chen, "Using Word Segmentation Model for Compression of Chinese Text"
<http:// nlplab.cs.nhtu.edu.tw/~mathis/own/html/PAPER/JNL/95/cpcol/ CPCOL95.htm>
Mathis H. C. Chen, Tsong-Yi Tseng, Jason J. S. Chang, "Automatic Generation of Indices or Chinese Books," <http://nlplab.cs.nthu.edu.tw/~mathis/own/html/ PAPER/JNL/96/cpcol/BookIdx.htm>