The ontology-driven approach to intelligent support of requirements engineering in agile software development

The paper presents the ontology-driven approach to intelligent support of requirements engineering while agile software development. An ontology system is proposed that combines the ontology for information support of the requirements engineering process and the ontology of the software product application domain. So we offer to analyze the requirements as logical statements about the application domain of the software product. The method of extracting and analyzing the requirements from natural Russian language formulations is described. The requirements are first processed by the UDPipe as a part of the method. Production rules for extracting ontology concepts are applied to the dependency trees built by UDPipe. The consistency of the requirements set presented in the form of the ontology instances is checked according to the developed production rules. The paper also gives a brief description of the developed decision support system (DSS) prototype which involves the joint use of several OWL files presenting knowledge about the project, the application domain and the semantic relations between the key elements of the proposals with the requirements (actors, actions and objects). The ontology editor Protege 5.2 is used to work with the ontology. The DSS is written in Python. DSS can exchange data with adjacent systems SWI-Prolog and UDPipe.