Large software development projects that use bug tracking systems can become overwhelmed by the number of reports filed.To assist in reducing the workload of project members, researchers have proposed the use of bug report assignment recommenders.To assist project members with the creation of assignment recommenders, we proposed a web-based tool called the Creation Assistant for Supporting Triage Recommenders (CASTR).This paper presents the results of both a laboratory and field study of CASTR.We found that CASTR can create assignment recommenders with accuracy as high as 95%, 80%, and 70% for Top-1, Top-3 and Top-5, respectively.The field study showed that 60% of the participants found CASTR easy to use, whereas the remaining participants found CASTR moderately or slightly easy to use.

Issue tracking systems are used to make a software development process more manageable, especially for a geographically dispersed team.However, for each bug report, a decision-making process called bug report triage needs to be performed.A common bug report triage decision is the assigning of a developer to a specific bug report.Bug report triage can take significant time and resources.Bug report assignment recommenders have been proposed for reducing the workload of a project member.However, creating a recommender is complex, as project members have to perform several steps such as data preparation and selection of a machine learning algorithm.Although previous work has sought to find specific answers for aspects of the assignment recommender creation process, to the best of our knowledge, only a few other works have examined assisting with this recommender creation process.CASTR (Creation Assistant for Supporting Triage Recommenders) [1, 2] is a platform-independent multi-tier web application.It allows a project member to analyze the dataset using a graphical representation and also assists in configuring project-specific parameters for a machine learning algorithm.This demonstration shows how to use CASTR to create a bug report assignment recommender, which consists of the following steps: