This application goal is training machine learning models using available online information, in GitHub and Jira, that a software testing team can exploit to discover the classes which more likely contain bugs, and focus the testing activities on these classes. Different optimal configuration will be searched for each project, actually the empirical study assesses the results on two open source project: avro and bookkeeper.
For more information, read the documentation