透過您的圖書館登入
IP:13.59.218.147
  • 學位論文

運用限制式校正距離與語意相似度比較模組化產品物料清單

Applying of Constrained Edit Distance with Semantic Similarity to Compare Bills of Materials of Modulized Products

指導教授 : 項衛中

摘要


在供應鏈環境中,企業持續與多方合作夥伴相互溝通和傳遞重要的商業訊息。在傳遞的資料方面,由於XML可供使用者妥善地表達資料並儲存,儼然成為大部分企業採用的資料交換格式。在新產品開發設計與製造過程中,設計者以XML資料格式設計模組化產品物料清單(bill of materials, BOM)並傳遞給製造商,但兩方卻可能因各自資料描述的方式與定義的不同而產生資料相異的問題。製造商若無法順利從現有的BOM資料庫中尋找出相似的設計來使用,將造成重複開發已存在的相似產品並增加額外的成本支出。 本研究期望開發一個XML綱要對映模組,以幫助企業從BOM資料庫中找出相似的模組化產品BOM來重複使用並設計,以避免重複開發的現象。在XML綱要對映模組中,採用限制式校正距離演算法作為綱要對映的核心技術,藉由演算法可解析出兩BOM同階層內結構相似性之特性,比較來源模組化產品BOM與目標模組化產品BOM之間的結構相似程度。再加入語意相關性的分析,透過WordNet語意詞庫查詢同義字求得語意相似度,以權重的方式加入結構對映中來提升綱要對映效能,並得到較佳的對映結果。 本研究利用範例驗證此XML綱要對映模組的功效。透過此對映模組比較兩份BOM的結構相似性,即使來源與或目標BOM結構在設計上有所差異,但在大多數的範例中除了可得到較佳的結構對映,亦可比較出來子結構之間的相似性。

並列摘要


Recently XML was recognized as a reliable data model to represent and store data; therefore, XML has emerged as a common data format for cross-platform information exchange between companies over the Internet in a supply chain environment. Product designer might use XML format to denote bill of materials (BOM) in designing and developing of new product, and delivered BOM to manufacturers. However, identical product data may be represented with different words or structures in individual information system. If manufacturers could find similar designs from their existing BOM databases, they will not re-develop existing components or products and waste production efforts. This research developed a XML schema matching module to find similar BOMs of modulized products from the BOM database for cooperated companies. This research basically applied the constrained edit distance method for unordered trees to solve the schema matching problem between source and target BOMs in XML schema format. This algorithm calculated the edit distance as the degree of structural difference with the restriction of comparing nodes in the same level of the BOMs. This method is useful to compare structural similarity between BOMs of modulized products. However, linguistic similarity based on WordNet database was included in this research to obtain weighted similarity for better matching results. This research uses several examples to verify the capability of the developed XML schema matching module. Based on the experimental results, even though BOMs with different structures, comparing the structural similarity between two BOM roots and their sub-structures may increase the matching accuracy and precision in most cases.

參考文獻


[21] 曾增偉,(2007),「包含共享綱要元件的XML文件交換問題的初探」,中原
[20] 劉冠宏,(2005),「運用校正距離計算結構相似度增進資料交換中綱要對映
[1] Bille Philip. (2004). A Survey on Tree Edit Distance and Related Problems.
[3] Lear A.C. (1999). XML seen as integral to application integration. IT Pro, 12-16.
[4] Lee Chiang, Chen Cgia-Juan, & Lu Hongjun. (1995). An aspect of query optimization in

被引用紀錄


劉凌宇(2010)。運用語意對映輔助電子化工作流程建置〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu201000745
高詠茹(2017)。從使用者觀點探討公共圖書館使用效益〔碩士論文,國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU201702446

延伸閱讀