自動化系統是圖書館和博物館最普遍的數位化工具之一，但每一個系 統均僅能處理單一的metadata，例如圖書館自動化系統的MARC。然而當各種的 知識需要數位化管理時，卻會產生許多不同知識型態與體系之間的整合與管理問 題。這些問題在國內博物館典藏品數位化時，便成了急需解決的一項工作。因此， 本研究提出一套系統架構與實作，能夠達成多種異質metadata並存處理的需求， 並透過檢索點的串聯定義，使各種異質metadata能夠在一致性的查詢功能之下 被檢索出來。而本系統的主要設計方法便是仿效XML的階層式架構，以及使用 DTD作為資料結構的綱要來源。

　　　　　Automatic information system is a common tool in the digitalization process in libraries and museums. Yet each system can handle only one single metadata, such as MARC for library system.When various data formats have to be digitalized in a complete system, there occur some problems in data integration and management, which become a key issue in the digitalization of collections in museums. If an information system is unable to process various formats of metadata, it lowers the efficiency and the utilization of metadata management. Therefore, we must design a novel system that can handle multi-metadata. This system is designed to build a hierarchical layout like XML, employing the DTD structure for its schema.