隨著知識經濟的發展，知識管理將越形重要，利用資訊科技的技術更可加速知識的運用與創造。結構化的資料比非結構化資料降低了知識探勘的選擇、處理及轉換資料的步驟，有助於知識管理系統萃取知識的品質與效率。
本論文主要研究是在知識著作者利用文書編輯器製作資訊文件的階段時，內嵌了資料片段組裝及文件樣式組裝的機制、結構化轉換的策略及領域屬性一致性檢查等功能，將供人們閱讀的非結構化資訊文件轉成結構化的資訊，並整合寫到企業組織的關聯式資料庫(RDB)中。如此可使得知識管理系統中的資料儲存處所趨向一致化，同時也可利用RDB應用在計算、統計、及資料分析等方面的優勢，來提升知識管理系統的效能。

Following by the development of knowledge economy, Knowledge Management is more important. The use of information technology will speed up the application and creation of the knowledge. Structured data is more reduced data mining selection, processing, and transformation data steps that can help knowledge management system to mining knows quality and efficiency than non-structured data.
This paper focuses on studying in knowledge author who used word processor editor to product information text. In this edit stage we had embedded data fragment and style-encapsulated functions, structural data transformation policy and domain attribution correspondence check functions. Unstructured text is easy to human readable. Transformation unstructured data to structured data and integrated to RDB of enterprise via this paper methods can make knowledge management system to store data and lead to correspondence, and can use RDB advantage likes to apply in calculation, statistic, graphic, and data analysis and so on, to prompt total knowledge management system efficiency.

一、緒論 1
1.1背景說明 1
1.2研究動機 2
1.3研究內容及方法 3
1.4論文架構 4
二、相關文獻及技術研究 5
2.1半結構化資料 5
2.2資料存取 10
2.2.1 ODBC介面 10
2.2.2利用個人電腦對資料庫進行存取 13
三、系統機制及策略分析 19
3.1資料片段組裝 19
3.2資訊萃取策略 21
3.3領域屬性一致性策略 25
四、系統架構及實作 27
4.1系統架構圖 27
4.2系統功能圖 30
五、系統成果討論 32
5.1討論 32
5.2系統展示 33
六、結論 40
6.1結論 40

[1] Ymashita, Rei-Jo., Ito, Tetsuro., Yao,H.,H., "On Management of Semi-Structured Data for Information system Development," SCI’99 and ISAS’99.
[2] Robert A. Nado., Scott B. Huffman., "Extracting Entity Profiles from Semi-structured Information Spaces," special section on management of semi-structured data, ACM SIGMOD RECORD, Dec. 1997.
[3] Kosko, B., "Bi-directional associative memories," IEEE Trans. on Systems, Man, and Cybernetic, 1987.
[4] Huffman, Scott., Steier, David., "Heuristic joins to integrate structured heterogeneous data," In Working notes of the AAAI Spring Symposium on Information Gathering in Heterogeneous Distributed Environments. AAAI, 1994.
[5] Huffman, Scott., Baudin, Catherine., "Notes Explorer: Entity-Based Retrieval in Shared," Semi-structured Information Spaces. In Proceedings of the 1996 International Conference on Information and Knowledge Management. 1996
[6]戴光志，「文件關聯式概念資訊萃取系統」，元智大學，碩士論文，民國八十八年。
[7]http://db.csie.ncu.edu.tw/DBMS/dblp/db/conf/pods/Buneman97.html
[8]http://dbweb.csie.ncu.edu.tw/DBMS/dblp/db/conf/vldb/GoldmanW97.html
[9]http://dbweb.csie.ncu.edu.tw/DBMS/dblp/db/conf/sigmod/NestorovAM98.html
[10]Ymashita,R.J., Yao,H.H. "ESSQL: An Enhanced Semi-Structured Query Language for Composite Document Retrievals", International Conference on SIGDOC, OCT 1998.
[11]P.Atzeni, G.Mecca, and P.Merialdo "Semi-structured and Structured Data in the Web: Going Back and Forth,” special section on management of semi-structured data, ACM SIGDOC RECORD, Dec 1997.
[12]J.McHugh, J.Widom, “Integrating Dynamically-Fetched External Information into a DBMS for Semi-structured Data,” special section on management of semi-structured data, ACM SIGDOC RECORD, Dec 1997.
[13]Yeh,L.J., Yao,H.H., Chen,Y.K. "SSQL: A Semi-structured Query Language for SGML Document Retrievals", International Conference on SIGDOC, OCT 1996.
[14]Dan Smith, Mauricio Lpoez, “Information Extraction for Semi-structured Documents”, GIE Dyade, INTRIA Ehone-Alpes, 1997.
[15]"Special Section on Management of Semi-structured Data", ACM SIGMOD RECORD, DEC 1997.
[16]姚修慎，Introduction to Database System，台北，國立編譯館，民
國八十七年。
[17]曾守正，資料庫系統之理論與實務，九版，儒林圖書有限公司，民國八十七年。
[18]Applehans, Globe, Laugero, Managing Knowledge-A Practical Web-Based Approach, Addison Wesley Longman, 1998.
[19]馮國扶，知識管理，跨世紀電子商務出版社，台北，民國八十八。
[20]馬曉雲，知識管理實務應用，微軟出版社，台北，民國八十九年。