本論文主要有二個研究方向，其一為分析各項機器學習理論，並將其應用於垃圾電子郵件的過濾器上，並同時支援中、英文的垃圾郵件分類。其二是針對郵件中各項特徵詞之間，研究如何去有效的考量各特徵詞之間的相互關係，提出改良式的稀疏二元多項式雜湊法，以達到更高的過濾準確率。
最主要的改良是應用於貝式的機率模型上進行分類，藉由應用稀疏的二元多項式雜湊法進行特徵值的前置處理，搭配適當的中文斷詞方法，提出改良過後的貝式模型，使其能應用於中文郵件問題之上，並將此分類器與其它分類器如支援向量機、KNN、K-Centroid等進行比較，同時建立一完整的Microsoft Access中文垃圾郵件語料庫，在Microsoft Windows平台上提出一個可行的中文郵件分類架構。
結合上述的改良式貝式模型，進一步藉由分析中文垃圾郵件的特性，提出一反饋主從式聯防架構，此系統結合MTA與MUA端的過濾機制，能發揮更強大的過濾功用，在阻隔竄改過之垃圾郵件，同時具備優異的過濾垃圾郵件之效能表現。

There are two major topics in which we engage, first, to analyze and apply each machine learning method in spam mails, which are written in English and Chinese words, classification problem. Second, according to the relation between features, we proposed a method, which combines "Sparse Binary Polynomial Hashing Method" with "Baysian Classifier." We believe this process could have better precision in classifying problem.
The main improvement is applying on Bayesian classifier. We combine this process to other machine learning methods. If we put in use advanced baysian method, we could get a better precision result. Most importantly, this procedure also works well in Spam written in Chinese. In the mean time, we build a well-define spam in Chinese database with Microsoft Access software, and then we could have a completely testing environment in modern mail system.
We join different process together and investigate the properties of modern spam mails. Finally, a new client-server system has been proposed. This system has better efficiency and more accurate precision. Based on above context, this system can help users build a clear, safely, and non-spam network environment.

[1] Andrew Troelsen: C# and the .Net Platform, 2nd. Apress Inc, 2003.
[2] Androutsopoulos, I., etc.: An experimental comparison of naive bayesian and keywordbased anit-spam ltering with personal email messages. In Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval, pages 160-167, 2000.
[3] Borenstein, N., Freed, N.: RFC Standard 1341, MIME (Multipurpose Internet Mail Extensions), 1992,
http://www.ietf.org/rfc/rfc1341.txt?number=1341
[4] Cohen, W., W.: Learning Rules that Classify E-Mail. In Proceeding of the AAAI Spring Symposium on Machine Learning in Information Access, 1996.
[5] Conference: Spam Conference In MIT. http://spamconference.org/, 2003-2005
[6] Drucker, H., Wu, D., Vapnik, V. N.: Support vector machines for spam categorization. IEEE Transactions On Neural Network, 1999.
[7] Graham, P.: A Plan for Spam, August 2002. http://paulgraham.com/spam.html
[8] Graham, P.: Better Bayesian Filtering. Proceedings of the 2003 Spam Conference, January 2003.
[9] Gregory, L., W., Wu, S., F.: On Attacking Statistical Spam Filters, First Conference on Email and Anti-Spam (CEAS), 2004.
[10] Han, E., H., Karmis, G.: Centroid-Based Document Classification:Analysis & Experimental Results, Computer Science Technical Report TR00-017, Department of Computer Science, University of Minnesota, Minneapolis, Minnesota, 2000.
[11] Han, E.-H. S., Karypis, G.: Centroid-based document classification: Analysis & experimental results. Tech. Rep. 00-017, Computer Science, University of Minnesota, Mar, 2000.
http://citeseer.ist.psu.edu/han00centroidbased.html
[12] Joachims, T.: Learning to Classify Text Using Support Vector Machines – Method, Theory and Algorithms. Kluwer Academic Publishers, Feb 2001.
[13] Johnson, K.: Internet Email Protocols – A Developer's Guide. Addison Wesley Longman Inc, 1999.
[14] Kaza, S., etc.: Identification of Deliberately Doctored Text Documents Using Frequent Keyword Chain (FKC) Model. Information Reuse and Integration, page(s):398-405, 2003. IRI 2003. IEEE International Conference, 27-29 Oct. 2003.
[15] Kotler, P.: Marketing Management. Prentice Hall, 2002.
[16] Lee, Y., J., Mangasarian, O., L.: SSVM: A Smooth Support Vector Machine for Classification, Computational Optimization and Applications, 20, 5-22, 2001.
[17] M. Crispin, M.: RFC Standard 2060, Internet Message Access Protocol, 1996, http://www.ietf.org/rfc/rfc2060.txt
[18] Meyer, T.A, Whateley, B.: SpamBayes: Effective open-source, Bayesian based, email classification system., First Conference on Email and Anti-Spam (CEAS), 2004.
[19] Michelakis, E., etc.: Filtron: A Learning-Based Anti-Spam Filter, First Conference on Email and Anti-Spam (CEAS), 2004.
[20] Myers, J., Rose, M.: RFC Standard 1939, Post Office Protocol - Version 3, 1996, http://www.ietf.org/rfc/rfc1939.txt?number=1939
[21] Pantel, P., and Lin, D.: SpamCop-- A Spam Classification & Organization Program. Proceedings of AAAI-98 Workshop on Learning for Text Categorization, 1998.
[22] Postel, J., B.: RFC Standard 821, Simple Mail Transfer Protocol, 1982, http://www.ietf.org/rfc/rfc0821.txt?number=821
[23] Radicati, S.: The Messaging Technology Report, In-Depth, Objective Analysis of the Messaging Industry. Volume 12, Number 8, August 2003. http://radicati.com/
[24] Rigoutsos, I., Huynh, T.,: Chung-Kwei: a Pattern-discovery-based System for the Automatic Identification of Unsolicited E-mail Messages (SPAM), First Conference on Email and Anti-Spam (CEAS), 2004.
[25] Sahami, M, etc.: A Bayesian Approach to Filtering Junk E-Mail. Papers from the AAAI Workshop, pp. 55–62, MadisonWisconsin. AAAI Technical Report WS-98-05, 1998.
[26] Salton, G., Buckley, C.: Term-weighting approaches in automatic text retrieval. In Information Processing and Management, pages 513-523, 1988.
[27] Tsai, C., H.: A Review of Chinese Word Lists Accessible on the Internet, 2004.
http://technology.chtsai.org/wordlist/
[28] Tsai, C., H.: MMSEG: A word identification system for Mandarin Chinese text based on two variants of the maximum matching algorithm, 2000.
http://technology.chtsai.org/mmseg/
[29] Vapnik, V., N.: The Nature of Statistical Learning Theory (Information Science and Statistics), Springer, 2 edition, November 19, 1999.
[30] Web Site: Fight Spam on the Internet! http://spam.abuse.net/
[31] Web Site: Open Relay Database. http://www.ordb.org
[32] Web Site: 中研院中文計算語言研究小組CKIP斷詞軟體1.0版, 2002. http://rocling.iis.sinica.edu.tw/CKIP/
[33] Witten, I., H., Eibe Frank: Data Mining – Practical Machine Learning Tools and Techniques with Java Implementations. Morgan Kaufmann, 2000.
[34] Woitaszek, M.,, etc.: Identifying Junk Electronic Mail in Microsoft Outlook with a Support Vector Machine. Applications and the Internet, page(s):27-31, Jan, 2003 Proceedings.
[35] Yerazunis, W., S.: Sparse Binary Polynomial Hashing and the CRM114 Discriminator, 2004. http://crm114.sourceforge.net/
[36] 侯捷: Word排版藝術, 碁峰圖書, 2002.
[37] 黃嘉輝: Visual Basic.Net 網際網路程式設計 – TCP/IP與Internet Programming, 文魁資訊, 2003.
[38] 謝居呈: Visual Basic.Net 程式設計指南, 新維文化科技, 2005.
[39] 吳昭逸: 具垃圾信過濾與安全機制之電子郵件收發系統, 台灣科技大學, 2004.