本研究的主要目標是使用名為XKD之XML型態之研究日誌來找出研究者的研究傾向(使用向量來描述) ，本系統可應用此XKD來自動建立研究者的研究領域的分類。如此分類的好處有兩點：(1)分類能夠讓無秩序的知識文件變成一個清晰有條理的的展現方式。(2)細分知識將幫助研究者將研究集中於小範圍當中。在分類的方法當中，我們的研究和其他人的研究是有所不同的。我們是針對研究者來作分類，而這是和其他研究的主要不同之處。本研究是從XML文件及XML文件的時間序來對研究者作分類，而一般的文件分類方法只是從文件的角度來作分類的動作。
我們提出了一個名為'WIDE R&D'的標準架構來完成研究者的資料交換動作。針對分散式群組的需求，我們也提出一個名為XKM的分類方法。此方法能夠在特徵的品質和研究者的時間進程之間取得平衡。最後，我們實作了這個演算法以及討論實驗的結果。

The main purpose of this research is to find out the researcher’s research-trend (is described by vectors) by XML-Specified research Logs called XKD. The system will apply XKD to build categories of researcher’s fields automatically. Advantages of classification have at least two points. (1) Classification will change from disorderly knowledge documents to clear presentations. (2) Subdivided knowledge will help for focus-research on the small scope. In general, our research is different with others classified methods. We aimed at the classification of researcher and that’s the main difference to others. Ordinary documents-classification methods only make the classifications from documents but our work can make classifications from researcher by XML documents and their time-order.
Our study proposed a standard framework named ‘WIDE R&D’ (Web-based, Integrated but distributed environment for research and design) to complete the exchange actions by distributed information of researchers. We aimed at researcher’s demand and proposed a method of categorization called ‘XKM’. The method can get balance between feature’s qualities and research’s time-progress. Finally, we had implemented the algorithm and discussed the results of experiment.

Abstract iv
Acknowledgement vi
Contents vii
Chapter1 Introduction 1
1.1 Motivation 1
1.2 Apply Scopes 2
1.3 Related research work 2
1.4 Our Problems 3
Chapter2 Background 5
2.1 KNOWLEDGE, KM and KMS 5
2.1.1 Explicit and Tacit Knowledge 7
2.2 XML protocol 8
2.2.1 Modeling Document 10
2.2.2 DTD Syntax 11
2.2.3 XML Schema 13
2.2.4 XML-RPC 15
2.3 DOM Module 16
Chapter 3 Data Mining 19
3.1 Association Analysis 22
3.2 Sequential relation mining 23
3.3 Clustering analysis 24
3.4 Classification analysis 27
3.4.1 Feature Selection 28
3.4.2 Traditional Classification Method (Decision Tree) 29
Chapter 4 System Framework 31
4.1 WIDE-R&D Framework 31
4.2 System functions & modules 32
4.2.1 Detail discuss 32
4.2.2 Modules in the system 33
4.3 XKD and XKDS 34
Chapter 5 Kernel of Classification 37
5.1 Overview 37
5.2 The method of XKM 38
5.2.1 Ctp (The Time Progress factor) 41
5.2.2 Get the characterization of one researcher 41
5.2.3 The characteristic of a category 42
5.2.4 Classification 43
5.2.5 Two Strategies in Classification 44
Chapter 6 Implementations and Experiments 45
6.1 System Overview and The Usage of XKDS System 45
6.2 Discussion 50
Chapter 7 Conclusion and future work 52
Reference 54

[1]W. Hsu, and S. Lang (1999), “NETNEWS Classification via Batch Routing and Updates”, Proceedings of Inter-national Conference of Information Resources Management Association.
[2]Wu-Xing Xie (1999), ”A knowledge management study aimed at ‘research papers’”, Thesis of National Chengchi University, Taiwan.
[3] Chengxiang Zhai and John Lafferty (2001), "Language Modeling Approach to Information Retrieval", ACM Conference on Information and Knowledge Management 2001, pp. 403-410
[4] Joachims, T.(1997), "A Probabilistic Analysis of the Rocchio Algorithm with TFIDF for Text Categorization", Proc. of the 14th International Conference on Machine Learning ICML97, pp. 143-151
[5] Bollacker, K.D. Lawrence, S. Giles, C.L (1998). CiteSeer: An Autonomous Web Agent for Automatic Retrieval and Identification of Interesting Publications, Proceedings of the Second International Conference on Autonomous Agents, Minneapolis MN, USA.
[6] Salton, G., and McGill, M.J.(1983), Introduction to Modern Information Retrieval, McGraw-Hill .Inc
[7] En-Hao Guan (1990), Fuzzy Correlation used in Text Multi-Categorization Problem, Thesis of Tamkang university, Taiwan.
[8] N. Cristianini and J.Shawe-Taylor(2000), An Introduction to Support Vector Machines. Cambridge University Press, Cambridge, UK.
[9] Senn, J. A. (1996), “Capitalizing on Electronic Commerce-The Role of the Internet in Electronic Markets”, Information Systems Management.
[10] A. Walsh (2002), UDDI, SOAP, and WSDL：The Web Services Specification Reference Book, PH.
[11] K. Duffey, R. Huss, V Goyal (2001), Professional JSP Site Design, WROX PRESS.
[12] Richard Monson-Haefel (2001), Java message Service, O’REILLY.
[13] Malhotra, Y. (2000), Knowledge Management for E-business Performance：Advancing Information strategy to Internet Time, Information strategy the executive’s journal, and Vol.16 (4), summer, pp. 5-16.
[14] Hong, T.& Han, I.(2002), “Knowledge based data mining of news information on the Internet using cognitive maps and neural networks”, Expert Systems with Applications., Vol. 23.
[15] Ian H. Witten and E. Frank (2000), Data Mining, Morgan Kaufmann, p.49