Title: An online paper reviewer recommendation method based on the combination of authority and activity

Authors: Hua Zhao; Li Wang; Qingtian Zeng; Wei Tao

Addresses: College of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao, Shandong, 266590, China ' College of Arts and Sciences Fundamentals, Qingdao Binhai University, Qingdao, Shandong, 266555, China ' College of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao, Shandong, 266590, China ' College of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao, Shandong, 266590, China

Abstract: The existing paper reviewer recommendation methods pay more attention to research interests, ignoring the integration of the reviewer's authority and activity. An online paper reviewer recommendation method combining authority and activity is proposed. Firstly, an expert citation network is established and the PageRank algorithm is adopted to evaluate expert authority. Secondly, a method for predicting the reviewer's domain activity based on time cycle is proposed. This method constructs an expert-keyword matrix with a time cycle at first, and then the matrix is smoothed, optimally decomposed and normalised to get the prediction matrix to be used to predict the reviewer's activity. Thirdly, a reviewer recommendation method that combines authority and activity is presented. Finally, experiments were carried out on the automatically collected data, and experimental results show that the proposed method is successful.

Keywords: paper reviewer recommendation; authority; activity; research interest; expert citation network; expert-keyword matrix.

DOI: 10.1504/IJCSM.2024.139921

International Journal of Computing Science and Mathematics, 2024 Vol.20 No.1, pp.46 - 57

Received: 20 Sep 2022
Accepted: 18 Jan 2024
Published online: 11 Jul 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article