Framework for automating requirement elicitation using a chatbot

Requirement elicitation determines the success of a project given that it describes the needs of a problem and its context at hand. However, a recent Chaos Report revealed that projects are increasingly failing in large numbers due to the poor quality of requirements resulting from an erroneous requirement elicitation phase (REP). This affects the scalability of the REP given the difficult nature of successfully executing it in complex environments. In this paper, we propose a framework to automate the REP, validating it with a system that consumes the IBM Watson Assistant (IWA) chatbot service. The validation is done by comparing traditional requirement elicitation with requirement elicitation using our system with the help of a Peruvian software company. Finally, we demonstrate that requirement elicitation with our implemented tool is more efficient than traditional requirement elicitation and results in higher quality requirements.