The Design and Development of an Online Quality Analysis System for Uml Structural Models in Software Reverse Engineering

Software reverse engineering (SRE) plays an important role in contemporary software developmentenvironment that emphasizes changing requirements and rapid development. SRE not only facilitates thedevelopment of software to generate the associated analysis and design contents after the software isimplemented, but also assists in rebuilding the analysis and design contents for existing legacy systems. Inobject-oriented (OO) system development with UML, a modeling technique for system’s analysis and design,existing research and studies aim at providing SRE solutions to convert a system’s code contents into UMLdiagrams based on specific OO languages. However, most of the studies focus on how precisely theconversion can reflect the system and there is little or no research that further looks into the quality of theconverted results. Since the review of software design like UML diagrams is knowledge intensive and seeingthe applicability of Ontology in organizational knowledge sharing, this research proposes to develop aknowledge-based ontological system for online quality analysis on UML structural models in SRE. In termsof structure and concept, the proposed system consists of a domain-specific knowledge and an inferenceengine. Specifically, the domain knowledge refers to the composition of quality definitions that focus onUML structural modeling diagrams and the inference engine comprises a rule base engine and the associatedinference design for diagnosing quality defects based on the input diagrams and providing relatedimprovement recommendations. Moreover, the system will allow users to provide feedbacks, in order tocontinually maintain and improve the system. To verify and validate the proposed work, this research willconduct an empirical case study involving a software development firm to demonstrate the real usage of thesystem and to investigate the benefits of the system bringing to the organization.