Improving Quality of elicited Requirements through Crowdsourcing: An Empirical Study

ABSTRACT The success of a software system mainly depends on the quality of its requirements. For general‐purpose software systems, traditional techniques of requirement elicitation often struggle to capture clear, complete, unambiguous and verifiable requirements, particularly when users are diverse and globally distributed. The traditional elicitation techniques do not guarantee high quality requirements. To address this limitation, this study investigates the use of crowdsourcing as a mechanism for requirements elicitation. Crowdsourcing uses a large group of online users and stakeholders who participate for scalable data collection, diverse user input and continuous feedback. This research work uses an exploratory mixed method approach conducted in two phases. In the first phase crowdsourcing is used for requirements elicitation and during the second phase, the quality of gathered requirements is assessed and improved through an online focus group. The result indicates an improvement in the quality of requirements elicited via crowdsouricng and improved via focus group i.e., the requirements have improved clarity, correctness, completeness, consistency, verifiability, traceability, and modifiability. The research work provides an empirical evidence of the improved quality of elicited requirements via crowdsourcing for general purpose software systems. It also offers practical insights to researchers and practitioners who are seeking to enhance requirement engineering practices.