Weighted P-Rank Algorithm Based on aHeterogeneous Scholarly Network

Jian Zhou,Lin Feng,Shenglan Liu,Jie Yang,Ning Cai

DOI: https://doi.org/10.21203/rs.3.rs-1080239/v1

2021-01-01

Abstract:The evaluation of scientific article has always been a very challenging task because of the dynamicchange of citation networks. Over the past decades, plenty of studies have been conducted on thistopic. However, most of the current methods do not consider the link weightings between differentnetworks, which might lead to biased article ranking results. To tackle this issue, we develop aweighted P-Rank algorithm based on a heterogeneous scholarly network for article ranking evaluation.In this study, the corresponding link weightings in heterogeneous scholarly network can be updatedby calculating citation relevance, authors’ contribution, and journals’ impact. To further boost theperformance, we also employ the time information of each article as a personalized PageRank vectorto balance the bias to earlier publications in the dynamic citation network. The experiments areconducted on three public datasets (arXiv, Cora, and MAG). The experimental results demonstratedthat weighted P-Rank algorithm significantly outperforms other ranking algorithms on arXiv andMAG datasets, while it achieves competitive performance on Cora dataset. Under different networkconfiguration conditions, it can be found that the best ranking result can be obtained by jointlyutilizing all kinds of weighted information.