An Approach to Identify Use Case Scenarios from Textual Requirements Specification
Saurabh Kumar Tiwari, Deepti Ameta, Asim Banerjee · 2019
Use case modeling refers to the process of identifying scenarios written in some natural language text, particularly to capture interactions between the system and associated actors. Several approaches have been proposed to maintain the synergy of use cases with other software models, but no systematic transformation approach is available to extract use case scenarios from the textual requirements specification. In this paper, we propose a systematic transformation approach that automatically extracts various use case elements from textual problem specifications. The approach uses Natural Language (NL) parser to identify Parts-Of-Speech (POS) tags, Type Dependencies (TDs) and semantic roles from the input text specification to populate use case elements. It further makes use of the questionnaire-based approach to develop the remaining unpopulated parts of the use case template. The paper demonstrates the applicability of the proposed approach by applying both industry and research-level case studies. The results highlight that the generated output is correct, consistent, non-redundant and complete, and helpful to use case developers in further analysis and documentation.