A Recommendation Approach to Diversify the Collaboration Scenario in Global Software Development Contexts
Tales Lopes, Victor Ströele, Regina M. M. Braga, José David, Fernanda Campos · 2021
Finding developers to assist with project issues is essential in Global Software Development (GSD) contexts, where various individuals with distinct characteristics are involved. Several recommending approaches lead to identifying the same group of individuals who end up work overloaded. Aiming to diversify the recommendation process, we introduced DRecSys, a diversity-based recommendation system. Our approach seeks to identify individuals with characteristics similar to those previously requested to collaborate. To that end, we proposed a hybrid process composed of supervised (classification) and unsupervised (clustering) techniques. We provided evidence that DRecSys is able to recommend suitable non-obvious developers to assist with project issues.