RStarRank: A Method for Identifying Rising Stars based on Academic Publication

Identifying rising stars in scientific fields is critical to many kinds of applications. Traditional talent evaluation methods are inapplicable for the recognition of rising stars. Moreover, existing recognition methods demonstrate some problems and limitations. This paper proposes a ranking-based method: RStarRank, which recognizes rising stars with machine learning algorithm. In other words, this method generates a prediction model through learning preliminary features of many beginners and their future H-index to predict sequential probability of young talents' growth into academic leaders. This paper takes the computer science subject for empirical study. The results show that, the prediction accuracy can reach over 90% for 10-year model and over 84% for 5-year model, RStarRank has been verified to have a certain validity and has a higher prediction accuracy than other similar algorithms. Finally, this paper summarizes and describes deficiencies of the proposed method.