Model-Based Design of User Story Using Named Entity Recognition (NER)

The needs analysis method to determine system specifications is generally represented as a user story. A user story is built to achieve the goals and values from the user’s point of view as the party requesting and will use the system. In this study, Named-Entity Recognition (NER) has been applied to build a model-based design for analysis techniques to identify, find and categorize user stories. Therefore, it can help system analysts and software development teams understand and improve user stories’ quality. The design model using spaCy’s library has an evaluation of the ten datasets. The order to evaluate the accuracy value is known for precision value is 97.45%, the recall value reached 99.67%, and the F1-Score was 84.61%. The experimental results show that the number of sentences does not affect the recall value and f1-score. However, the amount of data does not determine the accuracy of the label prediction.