Evaluating a tool based on historical data to assist risk management: a case study in software projects

A risk is an uncertain event or condition that, if it occurs, affects the objectives of projects. Risk management is costly and error-prone because risks are abstract and subjective. This paper aims to present a tool to assist in risk management in software projects. The methodological procedure adopted was a case study, collecting information through focus groups. We conducted experiments with real teams in software projects to evaluate the tool. In addition, to verify professionals' perceptions, we applied a questionnaire based on the TAM methodology. The Risk Control tool aims to make risk management more objective and systematic, reducing subjectivity in decision-making. The results pointed to the approach's usefulness in identifying and monitoring risks. However, professionals made reservations about the applicability of risk measurement. The tool's contributions are probabilistic inferences using Bayesian Network, which offers adapted responses to new inputs as soon as someone introduces them.