Towards Team Formation in Software Development: A Case Study of Moodle

Abstract

Software development is a team-based intensive activity where various skills (e.g. technical and analysis skills) are required to deliver high quality outcomes. An effective team member assignment is thus a crucial process. In this paper, we propose to adopt the existing machine learning approach for team recommendation to recommend software team members who are suitable for a given task. The approach take both individual strength and collaborative efficiency among team members into account to give a recommendation. We evaluate the approach on the Moodle project, well-known open source software project. The evaluation results show that the adopted approach yields a better recommendation performance compared to the baseline (i.e. random assignment approach).