Measurement of Similarity between Use Case Description and Sequence Diagram in Software Requirement Specification using Text Analysis for Dtrain Application
Rifqi Zaidan Irsyad Yanis, Yudi Priyadi, Shinta Yulia Puspitasari
Abstract
Dtrain application is built by implementing software documents, one of which is the Software Requirement Specification (SRS). In compiling software documents, there is a mismatch between one document element with other elements. Therefore, this study will measure the similarity between the Use Case Description (UCD) on the SRS case study Dtrain with Sequence Diagrams (SD) to determine how similar or compatible the two SRS elements are. This study aims to measure the similarity of UML artifacts between UCD and SD contained in a software development document. The suitability of the two artifacts can be done by analyzing the text in the description (UCD) and object (SD) sections. Based on the results and discussion, this study resulted in the following. First, a study that finds level of similarity between UCD and the SD by comparing the step results performed an extraction on the UCD and the extraction of messages and actors on the SD. Second, there are five use cases in the Dtrain application, so the UCD and SD are each 5. Third, the highest similarity value results are found in SD 1 and SD 5, with a similarity value of 0.712. Fourth, the results of the validation value using algorithm code to get a kappa score of 0.331. Fifth, the results of the validation obtained from results of the expert questionnaire got a value of 0.8905.