Data and Python scripts for "Improvement of Intervention Information Detection for Automated Clinical Literature Screening during Systematic Review"

Published: 8 September 2022| Version 1 | DOI: 10.17632/ccfnn3jb2x.1
Contributor:
Tadashi Tsubota

Description

Data and Python scripts for "Improvement of Intervention Information Detection for Automated Clinical Literature Screening during Systematic Review"

Files

Steps to reproduce

=== Prepare environment === 1. Just run the following: `pip install -r requirements.txt` === Prepare EBM-NLP original data === 1. Access the original repository and build data according to its README (run `build_data.py`). 2. Move `p2_i_dev.txt`, `p2_i_gold.txt` and `p2_i_train.txt` to `./EBM-NLP` === Set your email address to use PubMed API === 1. Open `./script/all_experiments.sh` and replace YOUR_EMAIL_ADDRESS with your own email address === Conduct all experiments === 1. Just run the following: `sh ./script/all_experiments.sh` 2. The results are stored in `./outputs/logs` as csv files.

Categories

Natural Language Processing

Licence