Group article recommendation based on ER rule in Scientific Social Networks

Group Recommendation Systems (GRS) is an emerging area in both research and practice and has been successfully developed in many domains as a type of information filter to overcome the information overload problem. With the growth of Scientific Social Networks (SSNs), the need for article recommendation is emerging. Considering that researchers can be grouped according to their research interests, and article recommendation to a group of users has not been addressed in the literature, this paper aims to develop and test an inferential model to accurately recommend articles for group researchers in SSNs. In this paper, a novel approach for group article recommendation, referred to as GPRAH_ER, is proposed to improve the processes of both individual prediction and group aggregation. In the stage of individual prediction, the Probabilistic Matrix Factorization method is adopted and is further unified by using articles' contents and group information. In the stage of group aggregation, the ER rule is introduced in the aggregation process, since it possesses the advantages of identifying group members' impacts based on the group member's weight and reliability. To verify the performance of the proposed method, experiments are conducted on a real dataset CiteULike. The experimental results show that the proposed GPRAH_ER method outperforms other benchmark methods, and provides a more effective recommendation of articles to researchers in SSNs. (C) 2021 Elsevier B.V. All rights reserved.