| المستخلص: |
The software development scheme starts with identifying the requirements for the analysis and starts the development processes. The requirements' process levels begin from analyzing the requirements to sketch the program's design, which is an essential task for programmers and software engineers. Also, many errors will occur during the requisite analysis cycle before transferring to other stages, which leads to a higher cost of the process than the initial particularized process. The motivation behind this is because of the designations of software requirements created in the natural language. To minimize these errors, we can transfer the software requirements to a computerized form using the UML diagram. To overcome the mentioned problem, a device has been produced. Plans can provide semi-automatized aid for designers to offer a UML class version from software program specifications using natural language processing techniques. The proposed method outlines the class diagram in a known configuration and, additionally, figures out the relationship between the instruction and the class diagram. In this research, we aim to enhance the procedure of producing the UML diagrams by utilizing the natural language, which will help the software developers analyze the software requirements with fewer errors and efficient way. The proposed approach will use the parser analysis and Part of Speech (POS) tagger to analyze the user requirements entered by the user in the English language. Next, select the verbs, phrases, etc. from the user's text. The obtained results confirmed that the proposed method got more reliable results than other techniques published in the literature. The proposed approach provides a better analysis of the given requirements and better diagrams, while also helping the software engineers.
|
