Automated screening of research studies for systematic reviews using study characteristics.

Journal: Systematic reviews
Published Date: Apr 25, 2018

Abstract

BACKGROUND: Screening candidate studies for inclusion in a systematic review is time-consuming when conducted manually. Automation tools could reduce the human effort devoted to screening. Existing methods use supervised machine learning which train classifiers to identify relevant words in the abstracts of candidate articles that have previously been labelled by a human reviewer for inclusion or exclusion. Such classifiers typically reduce the number of abstracts requiring manual screening by about 50%.

Authors

  • Guy Tsafnat
    Centre for Health Informatics, Australian Institute of Health Innovation, The University of New South Wales, Sydney, NSW 2052, Australia.
  • Paul Glasziou
    Centre for Research on Evidence Based Practice, Bond University, 14 University Drive (Off Cottesloe Drive), Robina, QLD 4226, Australia.
  • George Karystianis
    Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, Australia.
  • Enrico Coiera
    1Australian Institute of Health Innovation, Macquarie University, Level 6 75 Talavera Rd, Sydney, NSW 2109 Australia.

Keywords

External Resources

Popular Topics
Recent Journals