Recommending experts using communication history
Alan Moraes, Eduardo Silva, Cleyton Carvalho da Trindade, Yuri Barbosa, Sílvio R. L. Meira · 2010
In distributed software development the communication is inefficient because of geographical and temporal distances, affecting the team's performance and awareness. The low level of awareness makes hard the task of finding the expert of a piece of source code, delaying the implementation whenever a developer needs help. To identify and to recommend the people with right knowledge to people in trouble during the implementation can improve the collaboration and awareness of the team because it can reduce the waiting time for an answer, since the expert can be contacted directly. In this paper we propose recommender system for expert location with the aim to reduce delays of finding the right person whenever somebody needs assistance during coding. Our approach uses the communication history of the project (the developer's mailing list) in addition to usual source code history. We also present results which show the practical potential of our approach.