實驗室成員‎ > ‎

陳冠銘

陳冠銘 Kung-Ming Chan

碩士論文 (2005)

以Ontology為基之資訊整合架構設計
Design of Ontology-Based Information Integration Framework

關鍵字 Keywords

資訊整合, 本體論(Ontology), 語意相似度比對
Ontology, Information Integration, Semantic Matching

摘要

摘要   由於客製化的需求及全球化的競爭趨勢,使得企業多以策略聯盟的方式整合企業間價值活動與資源,以快速反應市場變化與滿足客戶需求。因此,「虛擬企業」被視為二十一世紀的主流商務模式。   在「虛擬企業」的運作環境中,使用到許多的IT技術,但由於參與的夥伴所使用的系統不同,因此企業間整合成為虛擬企業的成功關鍵。企業間整合可分為三個階段:資訊整合、系統整合與流程整合。資訊整合為企業間整合的基本工程,將會影響系統整合與流程整合,甚至影響虛擬企業的成敗。   本研究根據虛擬企業的特性與運作模式,分析現行的資訊整合架構,設計資訊整合架構所需的功能模組與軟體元件,並提出一Onto logy為基之Metadata與Metadata建構方法。本研究利用Ontology具有語意描述的特性及語意相似度比對的技術,來處理虛擬企業運作中資訊異動等問題,以協助改善現行資訊整合中語意異質性和資訊異動處理等問題。接著以物件導向模式化的方法,設計資訊整合平台之核心元件,並以Java開發軟體元件和採用Web Service的架構來部署網路元件,使軟體元件具有跨平台與分散性等特性。

Abstract

Abstract  Due to the demand of customization and the trend of globalization, enterprises are required to integrate value-added activities and resources across enterprises through enterprise alliance to quickly response the market changes and customers’ demand.”Virtual Enterprise” is consequently regraded as the most promising business model in the 21st century.  Information technologies are widely used in virtual enterprise. However, due to system heterogeneity, enterprise integration becomes a key factor for the success of Virtual Enterprise. Enterprise integration mainly includes three steps:Information Integration, Application Integration and Process Integration. Information integration is the fundamental construction of enterprise integration,due to it will affect application integration,process integration and the effectiveness of Virtual Enterprise.  In this study, the analysis and design of the information integration platform’s function modules and software components are first presented based on the characteristics and operational model of virtual enterprise and current information integration framework. Subsequently, the design of metadata and the development of the approach for metadata construction are conducted according to the characteristics of ontology. In this study, the uses of ontology matching and the technology of semantic similarity comparasion are employed to solve the problem of semantic heterogeneity and information variation.  Then, the design of core software using object-oriented method is presented. Finally, Java and web service are employed to develop software components, that makes software components have characteristics of distributability and integratability across platforms.


Comments