Research summary
Intelligent systems are technologically advanced machines that learn and respond to the world around them. They can take many forms—from automated vacuums to facial recognition programs to autonomous cars. But because these systems are trained to learn from their environment, their quality depends on the quality of the data used to train them.
Since traditional quality assurance (QA) does not consider these data, QA of intelligent systems is challenging for software engineers. Crucial new software quality attributes, such as fairness, are also not well-supported by existing QA methodologies. As Canada Research Chair in Quality Assurance of Intelligent Systems, Dr. Cor-Paul Bezemer aims to overcome these challenges. To do this, he and his research team are coming up with new techniques and tools for reporting and reproducing issues and developing better testing methodologies for intelligent systems.