Balanced Teams Formation using Hybrid Graph Convolution Networks and MILP

In this paper, we propose a novel model that is based on a hybrid paradigm composed of graph convolution network and Integer Programming solver. The model utilizes the potential of graph neural networks that have the ability to capture complex relationships and preferences among nodes. As the graph neural network forms node embeddings that are fed as input to the next layer of the model, the introduced MILP solver works to solve the team formation problem. Eventually, the experimental work shows that the outcome of the model is balanced teams.