地形圖徵鏈結資料架構

Framework of Linked Data for The Topographic Feature

地形圖徵為描述現實世界各類現象之成果，參考共同分類及語意架構、並以物件為基礎建立，具有空間範圍與位置、屬性描述、識別性考量之特性，可突破傳統參考底圖檢視角色之限制。如可將地形圖徵提供為國內各業務資料之一致性空間參考，可達到減少資源重複建置及提昇跨單位資料整合之效益。過去之流通作業多以資料集（Dataset）為單元，透過網路服務或API之公開介面，允許使用者介接使用。鏈結資料（Linked Data）近年之發展為網際網路流通及開放資料服務之模式帶來新的刺激，地形圖徵涵蓋全國範圍，其坐標及屬性由國家測製維護，可作為空間參考來源，因其資料品質穩定且持續更新，可為智慧國土之跨領域應用提供空間化及更新性之契機，鏈結資料思維之導入將可為地形圖徵之應用帶來新一代的運作模式。本研究提出以地形圖徵為核心之跨域鏈結資料基本架構與運作模式，以具有自我描述能力之地形圖徵為基礎，提出地形圖徵以資源描述架構（RDF）建立之具體規劃，再透過識別性之考量而串接不同領域之資料，由此建構位於網際網路之開放圖徵資料應用環境，提升探索跨域資訊之可能性，開創更為多元之應用契機。

The topographic feature is a critical media to describe the various phenomena in the real world. It refers to the common semantic classification of topographic feature, which is modeled as an object-oriented framework. The topographic feature has the characteristics of location, attribute, and identification. It is applied to many aspects of services, not just viewing. If the topographic feature can be provided as a consistent spatial reference for domestic domains, it can reduce the redundant construction of resources and improve the efficiency of cross-domain data integration. In the past, most of the operations used Dataset as a unit, allowing users to interface and use it through the open interface of web services or APIs. The development of Linked Data in recent years has brought new model to the web applications and open data services. The topographic feature has the advantages of nationwide spatial reference and continuous updating, which can provide cross-domain applications for smart land projects. The adaption of linked data will bring a new generation of using of topographic feature. This research proposes the basic rules and the framework of linked data for topographic feature. Topographic feature is recorded with the format of resource description framework (RDF), and has the ability of self-describing. With the design of identifier URI, topographic features can link to each other, even other cross-domain resources. This framework can enhance the possibility of exploring cross-domain data, and creating more diverse value-added application.