運用本體論整合資料庫主檔資料之研究

近年來資訊科學領域中，資料庫領域、語意網領域和電子科學期刊的許多研究與文獻，經常出現「以本體論為基礎的資料整合」、「語意存取資訊」、「以本體論為基礎的資料管理」與「資料整合」等關鍵字詞，這也反映學術界與產業界目前正積極發展以語意為導向的資料存取與整合之應用，以提高企業的彈性與簡化企業流程。若將本體論(Ontology)整合應用到資料中，將使資料容易分享與明確規範，並使機器能夠理解，如此達到快速資料整合，獲得單一事實的資料，並提高企業彈性與簡化流程，避免資訊錯誤情形發生。本研究發展一套以本體論為基礎之資料整合模式，以協助企業進行主檔資料管理，並以逆向工程為基礎來發展轉換關聯式資料表為本體論之演算法，透過分析資料庫綱要，來確定關聯式資料表轉換為本體論之轉換規則，關聯式資料表轉換為本體論之後，透過本體論合併工具「Prompt」將轉換後的本體論對齊與合併，而此合併後的本體論可透過使用者輸入本體論查詢語言「SPARQL」來產生有用資訊。本研究將此模式套用在主檔資料管理中，因主檔資料為企業常用且最核心之關鍵資料，但主檔資料不是分散於異質系統中，即是被大量儲存於資料倉儲中，難以有效利用。因此將此模式導入主檔資料管理中，並以本體論形式定義與表示主檔資料，以有效整合與管理主檔資料。

關鍵字

本體論；資料整合；逆向工程；主檔資料管理

並列摘要

In recent years, the field of information science, database and literature of Semantic Web, e-science journals that often appear the keyword of ontology-based, data integration, master data management and so on. It is means the academia and industry is developing a semantic-oriented data management and data integration, if we use the ontology in data integration , that would be make information sharing easy, specification clearly and make the machine understandable. We would to achieve data integration efficient, information access accurately and to avoid error with information. There are so many subsidiaries in large enterprises or multinational corporations, including information systems independent of each other, there are own database respective, the enterprise resource planning system just only integrate the information for internal business units. But with the business expanding and needs increasing, providing data in holistic and single factual views to large enterprises can be avoided error that called "garbage in, garbage out".This paper designs a ontology-based data integration model, and developes transformation rules algorithm that use reverse engineering to transform relational table to ontology. After transforming, we use the tool called “Prompt” to ontology alignment and merging. Before merging,we use ontology query language “SPARQL” to query and generate useful information. Our paper use the ontology-based data management mode in the master data, because it is the most core data and be used commonly in the organization. The ontology-based data integration in master data not only makes data defined explicitly and formalization, but also integrates, shares and manages master data effectively.

並列關鍵字

Ontology ； Data Integration ； Reverse Engineering ； Master Data Management

參考文獻

[1] 陳怡燕，以本體論為基之知識整合機制研發，碩士論文，國立成功大學製造工程研究所，2006年。

[3] Alalwan, N. A., “Ontological Approach for Database Integration”, De Montfort University, http://www.dora.dmu.ac.uk, 2011.

[4] Astrova, I., “Reverse engineering of relational databases to ontologies”, lecture notes in computer science, Springer, Vol. 3053, pp. 327–341, 2004.

[5] Brunner, J-S., Ma, L., Wang, C., Zhang, L., Wolfson, D. C., Pan, Y., and Srinivas , K., “Explorations in the use of semantic web technologies for product information management”, Conference on World Wide Web, 2007.

[6] Berners-Lee, T., Hendler, J., and Lassila, O., “The semantic web”, Scientific American, Vol. 284, No.5, pp. 28-43 , 2001.

國際替代計量

運用本體論整合資料庫主檔資料之研究

全文下載

主題瀏覽