Sentiment Analysis of Tweets Mentioning Research Articles in Medicine and Psychiatry Disciplines
Recently altmetrics (short for alternative metrics) are gaining popularity among researchers to identify the impact of scholarly publications among the general public. Although altmetrics have been widely used nowadays, there has been a limited number of studies analyzing users’ sentiments towards these scholarly publications on social media platforms. In this paper, we analyzed and compared user sentiments (positive, negative and neutral) towards scholarly publications in Medicine and Psychiatry domains by analyzing user-generated content (tweets) on Twitter. We explored various machine learning algorithms, and constructed the best model with Support Vector Machine (SVM) which gave an accuracy of 91.6%.
KeywordsAltmetrics Twitter Medicine Psychiatry Sentiment analysis Text classification
- 1.Alternative Metrics Initiative Phase 1 White Paper. National Information Standards Organization (NISO), Baltimore. http://www.niso.org/apps/group_public/download.php/13809/Altmetrics_project_phase1_white_paper.pdf
- 2.Arredondo, L.: A Study of Altmetrics Using Sentiment Analysis (2018). http://commons.lib.niu.edu/handle/10843/17878
- 4.Sesagiri Raamkumar, A., Ganesan, S., Jothiramalingam, K., Selva, M.K., Erdt, M., Theng, Y.-L.: Investigating the characteristics and research impact of sentiments in tweets with links to computer science research papers. In: Dobreva, M., Hinze, A., Žumer, M. (eds.) ICADL 2018. LNCS, vol. 11279, pp. 71–82. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-04257-8_7CrossRefGoogle Scholar