為了提供一個全方位之文件資料庫的資料萃取雛型系統,以滿足使用者各式各樣的查 詢模式,及符合目前公司行號使用數最廣泛的資料庫系統為著眼。本論文將以關聯式資料 模式為基礎,運用功能式資料模型(Functional Data Model)來組織鏈結(links) ,並透過此種鏈結屬性的建立,達到如同HyperLink之方式,來巡覽(navigation)其 相關之文件。 本論文中提出類似結構化查詢語言(Structured Query Language,SQL)之查 詢指令語法架構(Semi -SQL)以及文件-字詞查詢語言(Document Term Query Language,DTQL),做為本論文所建立之系統雛型的查詢語言。利用Pivot-Matrix 的分類概念,提昇索引技術與簽名檔技術,對文件資料庫存取路徑之設計與選擇,亦利用 EERT模式(Enhanced Entity Relationship Term Model)以擴充文件資料庫的 查詢能力,同時對所提出之方法論的查詢語意表達能力等策略問題亦提出初步之探討。
The main objective of this thesis is to provide an integrated extracting prototype system for document databases so that we can satify the user various query models to correspond with DBMS of updated companies. This thesis will use a relational data model as a basis , and apply the Functional Data Model (FDM) to organize the links between documents in order to navigate the related document as "Hyperlink". This thesis proposes a Semi-Structured Query Language (Semi-SQL) and Document Term Query Language (DTQL) , and use the Pivot-Matrix concept to improve the index skill and signature file technology. By using the Enhanced Entity Relationship Term model (EERT-model) , it will enlarge the query ability of the document database system. Besides , We also probe the query semantic expression ability for these methodology.