Extraction of Activity Diagrams Based on Steps Performed in Use Case Description Using Text Mining (Case Study: SRS Myoffice Application)
Reynaldi Prama Octavially, Yudi Priyadi, Sri Widowati
Abstract
In this research, extraction is carried out to change the text on an artifact to be processed through Text Mining to process and analyze the text whose results are validated, using the coefficient equation formula. There is a discrepancy between the flow in the Activity Diagram and the steps performed in the Use Case Description. Based on these problems, this study aims to extract an artifact to find similarities between the Activity Diagram and the actions taken in the Use Case Description and perform text analysis using the Text Mining method to obtain similarities, then validation and reliability are carried out. The extraction process is broken down into several documents according to the artifacts used, for Use Case Description, namely: UC01, UC02, UC03, UC04, and UC05. As for the Activity Diagram, namely: AD1, AD2, AD3, AD4, and AD5. The results obtained from this study are the highest similarity between activity diagrams and use case descriptions in sentences UC04 and AD4 with a total similarity of 0.51394086. The highest similarity between words lies in the words “request” and “application” with a complete similarity of 0.9231. Kappa's score with the Gwet's AC1 formula using Python programming is “Less than the Opportunity Agreement” with a number of −008583, while the questionnaire filled out by the Expert is “Almost perfect” with a score of 0.92246.