Automated systematic reviews by using Deep Learning and Active Learning
Branch: master
Clone or download
Latest commit d0f9904 Jan 22, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data First explorative work on ASR Jan 9, 2019
hpc Changing folder structures Jan 16, 2019
preprocessing First explorative work on ASR Jan 9, 2019
src Changing folder structures Jan 16, 2019
.gitignore First explorative work on ASR Jan 9, 2019
LICENSE Create MIT license (#1) Jan 14, 2019
README.md Add media coverage Jan 22, 2019
requirements.txt First explorative work on ASR Jan 9, 2019

README.md

Automated Systematic Review

This project is work in progress and not production ready.

Systematic Reviews are “top of the bill” in research. The number of systematic reviews published by researchers increases year after year. But performing a sound systematic review is a time-consuming and sometimes boring task. Our software is designed to take over the step of screening abstracts and titles with a minimum of papers to be read by a human (in the training set and in the final included set) and with zero false negatives (or any other small number).

Dutch newspaper NRC on this project "Software vist de beste artikelen uit een bibliotheek van duizenden."

Technical documentation

Follows.

Planning

  • Autumn 2018 - Prototype and test on some labelled data sets.
  • Winter 2019 – Finalize prototype with active learning + test on some labelled data sets.
  • Spring 2019 – Compare with other software and test with more labelled data sets
  • Autumn/Winter 2019 – Test with unlabeled data and compare to human performance.

Contact and contributors.

This project is part of the research work conducted by the Department of Methodology & Statistics, Faculty of Social and Behavioral Sciences, Utrecht University, The Netherlands.

For any questions or remarks, please contact Prof. Dr. Rens van de Schoot (a.g.j.vandeschoot@uu.nl).

Contributors: