Composing Scientific Collaborations Based on Scholars' Rank in Hypergraph

摘要

Finding the right scholars for collaboration is crucial for scientific progress. In this study, a novel algorithm is proposed to find the successful team configurations for scientific collaboration in the presence of the collaboration network of scholars. In this algorithm, the collaboration network is exploited to estimate the trust level among team members and the skill level of the scholars, while a hypergraph is used to model the relations. Also, our algorithm improves the search process by directing it to the promising regions, where the probability of finding the successful teams is high. A comparison with other algorithms is done to evaluate the proposed algorithm, using the similarity to successful collaborations. Our findings show that this algorithm achieves a significantly higher performance, compared to the other algorithms.