資訊管理學報
陳仲儼 Chung-Yang Chen,徐子涵 Tzu-Han Hsu;
頁: 379-412
日期: 2019/10
摘要: 軟體反向工程在現今講求快速實作與需求變化的軟體發展情境下越發受到重 視。以物件導向資訊系統發展來說，現有軟體反向工程研究皆針對特定的物件程 式語言以產生與系統內容相符的 UML（統一塑模語言）的系統模型圖、並發展自 動化工具。然而，這些研究大多是研發如何將程式碼準確地轉換成 UML 圖，鮮 少是針對圖形內容進一步檢核品質。此外由於軟體驗證如 UML 圖之品質檢核是 知識密集的工作，其需要來自組織經驗之不斷累積以使品質檢核更臻完善且與時 俱進。為此，本研究運用本體論、並針對系統設計品質中重要的耦合度量來發展 一套知識導向的軟體反向工程 UML 線上品質檢測系統，簡稱為 OntREDQ。在架 構與概念上，OntREDQ 由領域知識、規則引擎與品質推論所構成。其中領域知識 是以 UML 結構模型中的類別圖與耦合度量為主之品質本體模型；而規則引擎與 品質推論包含推論規則之設計以及針對圖形內容依據品質本體來分析與推論品質 缺失並提供改善建議。本研究所發展的工具僅為一研究雛形，但希望藉由雛型與 創新概念拋磚引玉，以供學術研究者或是發展 CASE 的實務企業所參考。
關鍵字: 軟體反向工程;本體模型;統一塑模語言;類別圖;耦合;
Abstract: Purpose－This research is to develop a knowledge-based ontological system, abbreviated as OntREDQ, for online quality analysis on UML structural models in SRE. Design/methodology/approach－In terms of structure and concept, OntREDQ comprises a domain-specific ontology that focuses on coupling and an inference engine to diagnose quality defects of the input diagrams and to provide improvement recommendations. Findings－The system is found to be able to effectively identify design coupling defects, as well as to allow users to provide feedbacks for continual improvement of the system. Research limitations/implications－As most of the SRE studies focus on how precisely the conversion can reflect the code, this paper further looks into the quality of the converted results. Practical implications－The proposed work helps project teams ensure design quality while jumping into implementation without system design under the SRE environment. Originality/value－This research contributes to developing a coupling measure that integrates the static and dynamic aspects, as well as to establishing an ontological application that is for the software design and reserve engineering domain.
Keywords: software reverse engineering, ontology, class diagrams, coupling;
瀏覽次數: 13132 下載次數: 1325
引用 導入Endnote
頁: 379-412
日期: 2019/10
摘要: 軟體反向工程在現今講求快速實作與需求變化的軟體發展情境下越發受到重 視。以物件導向資訊系統發展來說，現有軟體反向工程研究皆針對特定的物件程 式語言以產生與系統內容相符的 UML（統一塑模語言）的系統模型圖、並發展自 動化工具。然而，這些研究大多是研發如何將程式碼準確地轉換成 UML 圖，鮮 少是針對圖形內容進一步檢核品質。此外由於軟體驗證如 UML 圖之品質檢核是 知識密集的工作，其需要來自組織經驗之不斷累積以使品質檢核更臻完善且與時 俱進。為此，本研究運用本體論、並針對系統設計品質中重要的耦合度量來發展 一套知識導向的軟體反向工程 UML 線上品質檢測系統，簡稱為 OntREDQ。在架 構與概念上，OntREDQ 由領域知識、規則引擎與品質推論所構成。其中領域知識 是以 UML 結構模型中的類別圖與耦合度量為主之品質本體模型；而規則引擎與 品質推論包含推論規則之設計以及針對圖形內容依據品質本體來分析與推論品質 缺失並提供改善建議。本研究所發展的工具僅為一研究雛形，但希望藉由雛型與 創新概念拋磚引玉，以供學術研究者或是發展 CASE 的實務企業所參考。
關鍵字: 軟體反向工程;本體模型;統一塑模語言;類別圖;耦合;
The Development of Online and Ontological Quality Evaluation on UML Structural Design in Software Reverse Engineering: With a Focus on Coupling
Abstract: Purpose－This research is to develop a knowledge-based ontological system, abbreviated as OntREDQ, for online quality analysis on UML structural models in SRE. Design/methodology/approach－In terms of structure and concept, OntREDQ comprises a domain-specific ontology that focuses on coupling and an inference engine to diagnose quality defects of the input diagrams and to provide improvement recommendations. Findings－The system is found to be able to effectively identify design coupling defects, as well as to allow users to provide feedbacks for continual improvement of the system. Research limitations/implications－As most of the SRE studies focus on how precisely the conversion can reflect the code, this paper further looks into the quality of the converted results. Practical implications－The proposed work helps project teams ensure design quality while jumping into implementation without system design under the SRE environment. Originality/value－This research contributes to developing a coupling measure that integrates the static and dynamic aspects, as well as to establishing an ontological application that is for the software design and reserve engineering domain.
Keywords: software reverse engineering, ontology, class diagrams, coupling;
瀏覽次數: 13132 下載次數: 1325
引用 導入Endnote