Automatic Generation Method of Airborne Display and Control System Requirement Domain Model Based on NLP

Domain modeling is a crucial step from natural language requirements to precise specifications, and an essential support for the development of automation system design tools. The existing domain model extraction methods are not accurate enough to be applied into specific fields. In this paper, we present a method for extracting the requirement domain model of airborne display and control system based on Natural Language Processing (NLP). Firstly, the domain model template is defined on the basis of the detailed study of the existing rules. Then in the requirement statement, the parse tree generated from Stanford Parser is utilized to preprocess the requirements for special symbols and conjunctions. Finally, we conduct the comparative experiment and the results indicate that the precision of domain model extraction is 20.01% higher than the existing approaches without preprocessing.