Predicting public health impact: Linking ResearchGate presence to Scopus performance through machine learning

Ramin Rahmani¹, Heidar Mokhtari², Abbas Doulani³, Mohammad Karim Saberi^*4

1. Occupational Health Engineering, School of Public Health, Hamadan University of Medical Sciences, Hamadan, Iran
2. Department of Knowledge and Information Science, Payame Noor University, Tehran, Iran
3. Information Science Department, faculty of Education and Psychology, Alzahra University, Tehran, Iran
4. Department of Nursing, Shirvan Faculty of Nursing, North Khorasan University of Medical Sciences, Bojnurd, Iran
   *Corresponding Author: Mohammad Karim Saberi; E-mail: mohamadsaberi@gmail.com

Abstract

Background

ResearchGate as a main scientific social medium and Scopus as a known citation database have main role in sharing research output among specialists in different disciplines.

Objective

This study aimed to evaluate the performances of Iranian researchers in occupational health field and correlate some related variables. It also used regression analysis as one of machine learning approaches for predicting researchers’ scientific performance.

Methods

This descriptive cross-sectional study was conducted in 2024 on ResearchGate and Scopus indicators of Iranian researchers in the Occupational Health Engineering affiliated in Iranian universities (n=213). Data were extracted from ResearchGate and Scopus and the researches’ demographic information was collected from Iranian Scientometrics Information Database in medicine. Results 149 researchers (70%) were active in ResearchGate. 144 researchers (96.6%) had RG scores with the mean rate of 11.70. in ResearchGate, they shared total 4,275 research items with the mean rate of 28.89 items per researcher. With total 24,235 citations, the mean rate of citations per paper was 169.48. Of them,143 (95.9%) had ResearchGate h-indexes with the mean rate of 5.38. In Scopus, 198 researchers (93%) had total 2,935 published documents in the database with mean rate of 14.82 documents per researcher. 186 researchers (87.3%) had total 18,749 citations with the mean rate of 100.80 citations and mean h-index amounted to 4.41. Researchers with more shared documents in ResearchGate had better performance in Scopus. Linear regression analysis showed that the researchers’ presence in ResearchGate can predict their citation counts (R2=.82, β=.911, p=.000) and h-indexes (R2=.83, β=.900, p<.001) in Scopus.

Conclusion

Iranian researchers in the Occupational Health Engineering field fairly use the capacities of ResearchGate for influencing their research output. However, their interactions in social media tools should be encouraged for more reach and influence of their scientific productions.

Keywords: ResearchGate – Scopus – Research Evaluation – Iran – Occupational Health Engineering