Online ontological quality assessment of converted UML class diagrams in SRE

Software reverse engineering (SRE) is receiving more attention due to the current trend for rapid software development, since this allows developers to implement systems without any detailed design process and then use SRE tools to generate the design content, such as unified modeling language (UML) diagrams. In extant SRE research, studies have mainly focused on how precisely the conversion can reflect the system. However, little research has investigated the quality of the converted results, which could be inherently affected by poorly written software code. Accordingly, a knowledge-based ontological system (OntoRESec) was developed in this study to enable the online quality assessment of converted UML structural design content. With a focus on class design, OntoRESec features a domain-specific ontology model with a rule-based inference engine to assess the reversed class diagrams in SRE. With regard to the domain of quality assessment, the system is scoped to security in class design, since this is a major concern in terms of design quality and should be incorporated into the automated SRE context. An illustrative case is presented that adopts a security framework (STRIDE) to implement and demonstrate the rule-based inference of quality assessment in OntoRESec. In addition, several system cases are used to evaluate the proposed work.