Capability-Based Team Formation in Researchers Social Networks

Teamwork and cooperative working are increasingly needed due to the growing complexity of scientific problems and highly specialized research projects. Although many studies have been conducted on the subject of team formation in social networks, the challenge of forming a team of experts at optimum cost while maintaining maximum skill fulfillment persists. The main objective in team formation problem is to assemble a team of experts to satisfy all of the technical skill requirements of a given project while minimizing the communication and personnel cost. Electing a good leader for the team leads to better organizing and management of the entire team and also helps to lower the communication cost. Therefore, in this study three algorithms are proposed to identify a team of experts and a leader. These algorithms select the best leader and a team with the minimum cost by pruning the communication graph, identifying the effective nodes and choosing candidates for leadership according to various criteria. Moreover, a new combinational cost function is defined based on the linear combination of the objectives to minimize the personnel and communication costs. The results of experiments on a DBLP data set reveals that these algorithms are faster and more effective compared to other algorithms. The obtained results are due to the omission of excessive nodes according to the skills of experts and the project's requirements, along with choosing the leader based on appropriate criteria.