Research And Application On Consistency Detection Of Ontology-Based Feature Model

Improving the reusability of software in the same domain has been considered as an efficientway to develop software in the form of industrialization. The accurate analysis and verification ofrequirements are the basis of achieving high-quality software reuse. Building feature model has beenadopted as a dominated product line driven approach to capture domain requirements. However, it isredundant and confusing to depict feature models and constrains between them because of the lack offormal description and also not capable for verification in these methods. Therefore, a consistencydetection method for feature model based on ontology is proposed, which has been implemented incode level. In this thesis, presented approach could not only descript the feature model formally toavoid ambiguity and enhance information sharing, but also verify the model with reasoning ability ofontology to guarantee the quality of the software in requirements phase. This approach will bebeneficial to semantic modeling of domain feature models and their verification.Firstly, this thesis summarizes the shortcomings of traditional feature models, analyzes thecharacteristics of ontology-based feature models and the role it plays, generalizes the internal andexternal research on feature model.Secondly, an ontology-based feature modeling method that defines the feature classes, relationsand constrains among them is clarified. And then a meta-model of the ontology-based feature modelsand the transformation from feature model to ontology are put forward.Thirdly, this thesis explains the way for reasoning and verification for ontology-based featuremodel and Jena rules for mutex and require constraints are proposed to verify the consistency offeature model instances.Finally, the modeling process of the feature model with ontology and its verification arediscussed systematically against the background of Integrated Modular Avionics(IMA) domain. Thiscase study indicates the validity and availability of the method.