Screening evidence for systematic reviews using a text-mining system: The RobotAnalyst

Activity: Talk or presentation › Invited talk › Research

Description

Screening numerous citations for a systematic review can be laborious and costly, with the risk of missing relevant items among the retrieved hits. We shall present the results of recent research, in collaboration with NICE, into how a modest number of decisions by the human analyst can, in a few cycles of interactions, guide a system into reliably distinguishing what should be included and what should be excluded for a review. We shall also show how techniques from text mining can be used to support analysts working on public health reviews with a complex, even fuzzy, evidence base.
Period21 May 2019
Event titlePublic Health Data Science Day
Event typeConference
LocationColindale, United KingdomShow on map
Degree of RecognitionNational

Keywords

  • data science
  • systematic reviews
  • text mining
  • public health
  • active learning
  • machine learning

Research Beacons, Institutes and Platforms

  • Manchester Institute of Biotechnology

Documents & Links