On Privacy and Utility while Improving Software Quality

Abstract

Software development produces large amounts of data both from the process, as well as the usage of the software product. Software engineering data science turns this data into actionable insights for improving software quality. However, the processing of this data can raise privacy concerns for organizations, which are obligated by law, regulations and polices, to protect personal and business sensitive data. Early data privacy studies in sub-disciplines of software engineering found that applying privacy algorithms often degraded the usefulness of data. Hence, there is a recognized need for finding a balance between privacy and utility. A survey of data privacy solutions for software engineering data was conducted. Overall, researchers found that a combination of data minimization and obfuscation of data, produced results with high levels of privacy while allowing data to remain useful.