臺灣博碩士論文加值系統

聚類方法 (Clustering Method) 已經被大量的應用在電腦視覺 (Computer Vision) 和圖形識別 (Pattern Recognition) 上，但是對於某些聚類問題以目前較常用的的聚類演算法(如：Fuzzy C-Means、 Possibilistic C-Mean)，並無法得到令人滿意的結果，所以在本論文中嘗試提出一個適應性聚類演算法，希望能夠針對不同的資料分佈情形而適應性的產生出適合該種分佈的聚類中心 (Cluster Center)。
此外隨著資訊的增加，文件的數量也在增加中。透過聚類演算法 (Clustering Algorithm) 的使用，可以在每個類別中找出一個具有代表性的特徵文件，再根據聚類的結果可以找到與這代表性文件相關的文件。因此本論文嘗試將所提之適應性聚類演算法應用在文件聚類上。

Clustering methods have been used extensively in computer vision and pattern recognition. But common use clustering algorithm (like Fuzzy C-Means, Possibilistic C-Means) can''t gain satisfactorily result of every problems. Therefore try at this thesis to hand in an adaptive clustering algorithm, would like to aim at the different data distribution to find out the suitable cluster centers.
With the increasing of information, the number of documents also increasing very fast. People don’t want to spend much time to read every document to gain the information what they want. There must be an efficient method to find out the information what they want. Clustering algorithm can be used to solve the question above. Using the clustering algorithm, we can find out the prototype document in each cluster. According to the result of using clustering algorithm to analysis the data, we also can find out documents that related with the prototype document. In this thesis, we try to apply our adaptive clustering algorithm at document clustering.
英文摘要 ii
誌 謝 iii
目錄 iv
表目錄 vi
圖目錄 vii
第一章 緒論 1
1.1 研究動機 1
1.2 研究目的 2
1.3 研究方法簡介 3
1.4 章節簡介 4
第二章 文獻回顧 5
2.1 分割式聚類演算法 (Partitional Clustering Algorithm) 5
2.2 階層式聚類演算法 (Hierarchical Clustering Algorithm) 8
2.3 綜合分析 11
第三章 研究方法 13
3.1 基本定義和運算 13
3.1.1 歐幾里德距離 13
3.1.2 資料點權重 14
3.2 適應性聚類演算法 17
3.3 範例 20
第四章 實驗方法 23
4.1 二維向量空間實驗資料集 23
4.2 文件聚類應用資料集概述 23
4.2.1 文件 (Documents) 24
4.2.2 檔案格式 (File Format) 24
4.2.3 文件內部標示 25
4.2.4 類別 (Categories) 28
4.3 實驗設計 30
4.3.1 常用字去除 30
4.3.2 字根還原 31
4.3.3 特徵字選取 (Feature Selection) 31
4.3.4 文件表示方式 32
4.3.4 文件聚類應用實驗流程 32
4.3.5 正確率評估 33
4.4 實驗結果 35
實驗一 35
實驗二 40
實驗三 43
實驗四 47
實驗五 51
實驗六 52
第五章 討論與未來研究方向 53
5.1 討論 53
5.2 未來研究方向 53
參考文獻 55
附錄一 Stop-List 57

Bezdek, J.L., Pattern Recognition with Fuzzy Objective Function Algorithm, Plenum Press, New York, 1981.
Davé, R. N., “Characterization and detection of noise of clustering”, Pattern Recognition Letter, vol. 12, no. 11, pages 657-664, 1991.
Griffith, A., Luckhurst, H. C., and Willet, P., “Using Inter-Document Similarity Information in Document Retrieval Systems”, Journal of American Society for Information Science, 37, pages 3-11, 1986.
Jain, A. K. and Dubes, R. C., Algorithms for Clustering Data, Prentice Hall, Eaglewood Cliffs, New Jersey, pages 55-142, 1988.
Jardine, N. and Rijsbergen, C. J., “The Use of Hierarchical Clustering in Information Retrieval”, Information Storage and Retrieval, 7, pages 217-240, 1971.
Krishnapuram, R. and Keller, J. M., “A Possibilistic Approach to Clustering”, IEEE Trans. Fuzzy Syst., vol. 1, pages 98-110, May 1993.
Kim, H. J. and Lee, S.G., “A Semi-Supervised Document Clustering Technique for Information Organization”, In Proceedings of the 9th International Conference On Information knowledge management, pages 30-37, 2000.
Krishnapuram, R. and Keller, J. M., “The Possibilistic C-Means Algorithm: Insights and Recommendations”, IEEE Trans. Fuzzy Syst., vol. 4, pages 385-393, August 1996.
Lewis, D. D., ”Representation and Learning in Information Retrieval”, Ph.D. thesis, Computer Science Dept., Univ. Massachusetts at Amherst, Technical report pages 91-93, Feb. 1992,
Lewis, D. D., “The Reuters-21578 Text Categorization Test Collection”, http://www.research.att.com/~lewis/reuters21578.html, 1996.
Porter, M.F., “An algorithm for suffix stripping”, Program, vol. 14, no. 3, pages 130-137, 1980.
Lloyd, S.P., “Least-square quantization in PCM”, Bell Laboratories Internal Technical Report, 1957.
Steinbach, M., Karypis, G., Kumar, V., “A Comparison of Document Clustering Techniques”, In Proceedings of World Text Mining Conference, KDD 2000.
Zamir, O., Etzioni, O., Madani, O., and Karp, R. M., ”Fast and Intuitive Clustering of Web Documents”, In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining, pages 287-290,1997.
洪鵬祥, 張智星, 張俊盛, “中文新聞自動群聚”, 碩士論文, 國立清華大學資訊工程所, 新竹, 2000.