Skill Recommendation for New Contributors in Open-Source Software

Selecting an appropriate task is challenging for newcomers to Open Source Software (OSS) projects. Therefore, researchers and OSS projects have proposed strategies to label tasks (a.k.a. issues). Several approaches relying on machine learning techniques, historical information, and textual analysis have been submitted. However, the results vary, and these approaches are still far from mainstream adoption, possibly because of a lack of good predictors. Inspired by previous research, we advocate that the prediction models might benefit from leveraging social metrics. In this research, we investigate how to assist the new con-tributors in finding a task when onboarding a new project. To achieve our goal, we predict the skills needed to solve an open issue by labeling them with the categories of APIs declared in the source code (API -domain labels) that should be updated or implemented. Starting from a case study using one project and an empirical experiment, we found the API -domain labels were relevant to select an issue for a contribution. In the sequence, we investigated employing interviews and a survey of what strategies maintainers the strategies believe the communities have to adopt to assist the new contributors in finding a task. We also studied how maintainers think about new contributors' strategies to pick a task. We found maintainers, frequent contributors, and new contributors diverge about the importance of the communities and new contributors' strategies. The ongoing research works in three directions: 1) general-ization of the approach, 2) Use of conversation data metrics for predictions, 3) Demonstration of the approach, and 4) Matching contributors and tasks skills. By addressing the lack of knowledge about the skills in tasks, we hope to assist new contributors in picking tasks with more confidence.