Dr. MH. Nekoofar

DDS, MSc, DoIBoE, PhD

Tag: Machine learning

  • Science Map of Cochrane Systematic Reviews Receiving the Most Altmetric Attention Score: A Network Analysis

    Abstract

    The present study aimed to analyze and visualize the science map of Cochrane systematic reviews (CSRs) with high Altmetric attention score (AAS). On 2020-07-29, the altmetric data of the Cochrane Database of Systematic Reviews were obtained from the Altmetric database (Altmetric LLP, London, UK). Bibliometric data of the top 5% AAS of CSRs were extracted from the Web of Science. Keyword co-occurrence, co-authorship   and   co-citation   network   analyses   were   then   employed   using   VOSviewer software. The random forest model was used to rank the importance of the altmetric resource. A total of 11222 CSRs with AAS were found (Total mentions: 305265), with Twitter being the most popular Altmetric resource. Consequently, the top 5% AAS (649 articles, mean AAS: 204.95, 95% confidence level: 18.95, mean citations:  123.68, 95% confidence level: 13.9) were included. Density mapping revealed female, adult and child as the most popular author keywords. According to network visualization, Helen V. Worthington (University of Manchester, Manchester, UK), the University of Oxford and UK had the greatest impact on the network at the author, organization and country levels respectively. AAS were weekly correlated with citations (rs=0.21) although citations were moderately correlated with policy document and blog mentions (rs=0.46 and rs=0.43). Cochrane systematic reviews received high levels of online attention, particularly in the Twittersphere and mostly from the UK. However, CSRs were rarely publicized and discussed using recently developed academic tools, such as F1000 prime, Publons and PubPeer.

    Keywords: Cochrane systematic review, Altmetric, Bibliometric, Twitter, Machine learning, Network analysis, Random forest

     Science Map of Cochrane Systematic Reviews Receiving the Most Altmetric Attention Score A Network Analysis.pdf

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

    Abstract

    Introduction: We aim to visualize and analyze the science map of Cochrane systematic reviews with the high altmetric attention scores. Methods: On 10 May 2019, altmetric data of Cochrane Database of Systematic Reviews obtained from Altmetric database (Altmetric LLP, London, UK). Bibliometric data of top 5% Cochrane systematic reviews further extracted from Web of Science. Keyword co-occurrence, co-authorship and co-citation network visualization were then employed using VOSviewer software. Decision tree and random forest model were used to analyze citations pattern.  Results: 12016 Cochrane systematic reviews with Altmetric attention are found (total mentions=259,968). Twitter was the most popular altmetric resource among these articles. Consequently, the top 5% (607 articles, mean altmetric score=171.2, Confidence Level (CL) 95%=14.4, mean citations= 42.1, CL 95%=1.3) with the highest Altmetric score are included in the study. Keyword co-occurrence network visualization showed female, adult and child as the most accurate keywords respectively. At author level, Helen V Worthington had the greatest impact on the network. At organization and country levels, University of Oxford and U.K had the greatest impact on the network in turn. Co-citation network analysis showed that Lancet and Cochrane database of systematic reviews had the most influence on the network. However, altmetric score do not correlate with citations (r=0.15) (Figure 7), it does correlate with policy document mentions (r=0.61). Results of decision tree and random forest model (a machine learning algorithm) confirmed importance of policy document mentions. Discussion: Despite popularity of Cochrane systematic reviews in Twittersphere, disappointingly, they rarely shared and discussed in newly emerging academic tools (e.g. F1000 prime, Publons and PubPeer). Overall, Wikipedia mentions were low among Cochrane systematic reviews, considering the established partnership between Wikipedia and Cochrane collaboration. Newly emerging and groundbreaking concepts, e.g. genomic medicine, nanotechnology, artificial intelligence not that admired among hot topics.

    Keywords: Cochrane Systematic Review, Altmetrics, Science Map, Twitter, Facebook, Social Media, Citation, Policy Document, Network Visualization, Decision Tree, Random Forest, Machine Learning