資訊管理學報
陳滄堯;戚玉樑;
頁: 445-474
日期: 2015/10
摘要: 本研究發展一個以知識本體為基礎的知識系統，用於解決「供應商關係網絡」的追溯問題。由於現代企業面臨時效、成本、創新等經營壓力，商業環境已朝向專業分工及團隊結盟，以成為具有競爭力的供應商網絡。然而，受限於管理工具及資訊不對稱問題，大多數企業不易掌握第一層以外的供應商資訊，使得企業對突發狀況的應變能力不足，許多的證據顯示：因上游缺料而導致生產停滯，已造成企業經營的損失，因此亟須改善這類連鎖效應導致的問題。本研究提出一個以知識分類及正規語意建構的「供應商關係網絡」模型，藉由建立遞迴機制提供企業具有「追溯」供應商關係的能力。本研究主要設計方法包括：（1）將構成供應商關係的概念、屬性，利用知識本體（ontology）建置為知識模型；（2）將追溯上游供應商的步驟，利用語意規則發展推理程序，協助推論隱含性知識。本研究以太陽能產業24家相關企業資料進行實驗，由結果顯示：知識模型定義問題解決的內涵及關係，因此可串聯已知事實，推論供應商的關係網絡，包括追溯上游各層的供應商、潛在供應商、及獲取供應商的附帶資訊。簡言之，本研究的設計提供易於實踐的知識系統，可做為解決供應商關係延伸問題的基礎。
關鍵字: 供應商關係管理;供應商關係網絡;追溯;知識本體;語意規則;
Abstract: Purpose-This study aims to develop an ontological knowledge-based system to address the issue of supplier tracing in the supplier relationship networks. Design/methodolgy/approach-We develop an ontological knowledge-based system (KBS) based on tree recursive algorithm. The design includes two major components: 1) Using OWL-based approach to construct the concepts and attributes of supplier relationship into the knowledge framework; 2) Creating SWRL-based rules to define the inference processes in terms of the steps of tracing upstream suppliers. The created inference mechanisms then enable the reasoning of implicit knowledge from the known facts. Findings-The experimental results show that the knowledge model has defined the content and relations of the problem domain and it is capable of, through chaining finite facts, inferring through the supplier relationship network for tracing suppliers at all levels. The inference results include knowledge such as potential qualified suppliers and associated suppliers. Research limitations/implications practical implications-To make the supplier relation network feasible for industry tracing, we have simplified the required information and excluded sensitive data when modeling. The logic of supplier network relationships is embedded in the design and the suppliers need only provide the known and public factual information to conduct the hierarchical relationship tracing. The production expansion problem implemented in this study can be used as a model to extend into a more comprehensive design. Originality/Values-Two major original methodological contributions are present in this study. The first is the use of tree structure to process the recursive algorithm for the analysis of foundational concepts. The second one is the development of OWL-based KBS. The OWL data model was developed to represent class structure and does not natively support KBS development.
Keywords: supplier relationship management;supplier relationship network;trace;ontology;semantic rules;
瀏覽次數: 27967 下載次數: 120
引用 導入Endnote
頁: 445-474
日期: 2015/10
摘要: 本研究發展一個以知識本體為基礎的知識系統，用於解決「供應商關係網絡」的追溯問題。由於現代企業面臨時效、成本、創新等經營壓力，商業環境已朝向專業分工及團隊結盟，以成為具有競爭力的供應商網絡。然而，受限於管理工具及資訊不對稱問題，大多數企業不易掌握第一層以外的供應商資訊，使得企業對突發狀況的應變能力不足，許多的證據顯示：因上游缺料而導致生產停滯，已造成企業經營的損失，因此亟須改善這類連鎖效應導致的問題。本研究提出一個以知識分類及正規語意建構的「供應商關係網絡」模型，藉由建立遞迴機制提供企業具有「追溯」供應商關係的能力。本研究主要設計方法包括：（1）將構成供應商關係的概念、屬性，利用知識本體（ontology）建置為知識模型；（2）將追溯上游供應商的步驟，利用語意規則發展推理程序，協助推論隱含性知識。本研究以太陽能產業24家相關企業資料進行實驗，由結果顯示：知識模型定義問題解決的內涵及關係，因此可串聯已知事實，推論供應商的關係網絡，包括追溯上游各層的供應商、潛在供應商、及獲取供應商的附帶資訊。簡言之，本研究的設計提供易於實踐的知識系統，可做為解決供應商關係延伸問題的基礎。
關鍵字: 供應商關係管理;供應商關係網絡;追溯;知識本體;語意規則;
A Traceable Supplier Relationship Network Knowledge-Based System based on Recursive Structure
Abstract: Purpose-This study aims to develop an ontological knowledge-based system to address the issue of supplier tracing in the supplier relationship networks. Design/methodolgy/approach-We develop an ontological knowledge-based system (KBS) based on tree recursive algorithm. The design includes two major components: 1) Using OWL-based approach to construct the concepts and attributes of supplier relationship into the knowledge framework; 2) Creating SWRL-based rules to define the inference processes in terms of the steps of tracing upstream suppliers. The created inference mechanisms then enable the reasoning of implicit knowledge from the known facts. Findings-The experimental results show that the knowledge model has defined the content and relations of the problem domain and it is capable of, through chaining finite facts, inferring through the supplier relationship network for tracing suppliers at all levels. The inference results include knowledge such as potential qualified suppliers and associated suppliers. Research limitations/implications practical implications-To make the supplier relation network feasible for industry tracing, we have simplified the required information and excluded sensitive data when modeling. The logic of supplier network relationships is embedded in the design and the suppliers need only provide the known and public factual information to conduct the hierarchical relationship tracing. The production expansion problem implemented in this study can be used as a model to extend into a more comprehensive design. Originality/Values-Two major original methodological contributions are present in this study. The first is the use of tree structure to process the recursive algorithm for the analysis of foundational concepts. The second one is the development of OWL-based KBS. The OWL data model was developed to represent class structure and does not natively support KBS development.
Keywords: supplier relationship management;supplier relationship network;trace;ontology;semantic rules;
瀏覽次數: 27967 下載次數: 120
引用 導入Endnote