Designing a System to Aid in Conversational Requirement Elicitation Practices : Integrating REConSum and Trace2Conv into a system that can be used and experienced in practice

The thesis outlines the development of REConSum, a prototype aiding Business Analysts in agile software development by utilizing natural language processing (NLP) to automatically process conversation data from elicitation interviews, facilitating structured requirement extraction. It integrates multiple subsystems, including MS Teams integration, NLP processing, GUI interface, and Marker for data filtering, streamlining user interaction. REConSum eliminates manual setup, providing a user-friendly GUI interface. Although requiring refinement for specific user needs, it serves as a valuable tool for Fizor to gather feedback and iterate on its development without disrupting existing workflows.