Software Development for Object Formation for Sequence Diagrams Based on Use Case Description Extraction: Case Study SRS Scenery

In designing software, developers need an SRS document containing system provisions or requirements. The SRS file should have related contents between artifacts based on the elicitation process. Use Case Description is the result of the elicitation process and used to calculate Similarity value with the Sequence Diagram. However, testing reveals an incompatibility issue between Use Case Description and Sequence Diagram. This can happen because there are differences in determining objects based on the Step Performed in the Use Case Description. The research focuses on Scenery, a mobile-based tourism information app in Indonesia. The purpose of this research is to develop software that performs the formation of all objects for Sequence Diagrams through the extraction of Text Mining in Use Case Descriptions. The result of this research activity is application development that starts with analyzing the Step Perform artifacts contained in the Use Case Description and Sequence Diagram. Then create an artifact model by labeling Step Perform dataset from d1-d15 and d16-d30 for the Sequence Diagram. Text preprocessing using Python and implementing Cosine Similarity and Cohen's Kappa for validity testing yield the highest object forming values: 0.820368, 0.789612, 0.778272, 0.778272, and 0.77754. These value categories are "Almost Perfect agreement" and "Substance agreement". Based on the results of the five highest similarity value data, the objects suitable for the Sequence Diagram are as follows: user, content creator, scenery app, home page, content page, profile page, notifications page, register page, authentication server, login page.