Prediction of Rising Stars from Pakistani Research Communities

Abstract

Finding the rising stars is an interesting problem and has been recently studied in various domains including academic networks. This study formulates the problem of rising stars prediction as a machine learning task. Classification models are applied and features are classified as co-authors, author, and venues. The impact of the categorization of these features is empirically analyzed. The experimentation is performed on data of Pakistani researchers retrieved from Web of Sciences ranging from 2008 to 2014. After collection of the dataset, feature sets are calculated and classification techniques are applied. The researchers predicted by the proposed method are compared with top researchers of Pakistan in 2016 and 2017. The proposed technique solves the problem of finding rising star researcher of Pakistan in the field of computer sciences based on authors’ contribution, mutual influence, and venue citations scores.