Science Map of Cochrane Systematic Reviews Receiving the Most Altmetric Attention: Network Visualization and Machine Learning Perspective

Abstract

We aimed to analyze and visualize the science map of Cochrane systematic reviews (CSR) with high Altmetric attention scores (AAS). On 10 May 2019, the Altmetric data of the CSR Database were obtained from the Altmetric database (Altmetric LLP, London, UK). Bibliometric data of the top 5% of CSR were extracted from the Web of Science. Keyword co-occurrence, co-authorship, and co-citation network analysis were then employed using VOSviewer software. A Random forest model was used to analyze the citation patterns. A total of 12016 CSR with AAS were found (Total mentions: 259968) with Twitter being the most popular Altmetric resource. Consequently, the top 5% (607 articles, mean AAS: 171.2, 95% confidence level (CL): 14.4, mean citations: 42.1, 95%CL: 1.3) with the highest AAS were included in the study. Keyword co-occurrence network analysis revealed female, adult, and child as the most popular keywords. Helen V. Worthington (University of Manchester, Manchester, UK), and the University of Oxford and UK had the greatest impact on the network at the author, organization and country levels respectively. The co-citation network analysis revealed that The Lancet and CSR database had the most influence on the network. However, AAS were not correlated with citations (r=0.15) although they were correlated with policy document mentions (r=0.61). The results of random forest model confirmed the importance of policy document mentions. Despite the popularity of CSR in the Twittersphere, disappointingly, they were rarely shared and discussed within the new academic tools that are emerging, such as F1000 prime, Publons, and PubPeer.