1 / 15

Automating Study Discovery: Enhancing Systematic Review Efficiency through Semi-Automation Techniques

This article discusses innovative methods for automating study discovery in systematic reviews, focusing on two primary approaches: a pipeline system that connects publication to review topics and a semi-automated screening process. Utilizing citation networks and advanced search strategies, the methods aim to prioritize studies and streamline the screening burden. By leveraging active learning and text mining, researchers can create a more efficient ordering of titles and abstracts, allowing for earlier identification of relevant studies and facilitating the review process.

tanisha-gilliam
Télécharger la présentation

Automating Study Discovery: Enhancing Systematic Review Efficiency through Semi-Automation Techniques

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Automating study discovery James ThomasEPPI-Centre, Social Science Research Unit, Institute of Education, University of London

  2. Automating study discovery; how? • Two methods • A ‘pipeline’ between publication and reviews / topics • Semi-automating screening: ‘priority screening’

  3. Method 1: the ‘pipline’

  4. Completed reviews • RG specialist registers • Search strategies • Citation networks… • …? Mrs. Beeton's Book of Household Management by Isabella Beeton

  5. Cochrane Evidence Pipeline • Completed reviews • RG specialist registers • Search strategies • Citation networks… • …?

  6. Method 2: semi-automating screening

  7. The screening process Read title & abstract Click include / exclude Move on to the next reference Repeat…

  8. Changing the distribution of studies Traditional screening Screening process (red = eligible study) Screening aided by text mining

  9. Screening prioritisation (via active learning)

  10. The result • The result is an ordered list of titles and abstracts • With those that are most similar to the ones already marked as ‘include’ at the top • The person screening continues to screen as usual, but behind the scenes the titles and abstracts remaining are re-ordered regularly (e.g. every 25 items)

  11. Does it work? Empirical studies

  12. Comparison: two reviews ‘Basic’ methods With enhancements “Micro nutrients” “Cooking skills” Miwa M, Thomas J, O’Mara-Eves A, Ananiadou S (2014) Reducing the screening burden in systematic reviews through active learning on imbalanced data sets. Journal of Biomedical Informatics

  13. Using the technology in “live” reviews: not all textbook graphs!

  14. Possible uses for the methods & tools • Screening prioritisation (‘new’ reviews) • Finding relevant studies earlier in the review process (allowing full text retrieval to begin) • Curtailing screening before every citation has been checked manually • Facilitates early discussion and clarification of exclusion criteria • Automatic classification / instant ranked list for review updates

  15. Thank you SSRU website: http://www.ioe.ac.uk/ssru/ SSRU's EPPI website: http://eppi.ioe.ac.uk Email j.thomas@ioe.ac.uk EPPI-Centre Social Science Research Unit Institute of Education University of London 18 Woburn Square London WC1H 0NR Tel +44 (0)20 7612 6397 Fax +44 (0)20 7612 6400 Email eppi@ioe.ac.uk Web eppi.ioe.ac.uk/ The EPPI-Centre is part of the Social Science Research Unit at the Institute of Education, University of London

More Related