- The folder classifiers contains pickle files representing the Account Type and Occupation Type classfiers, which can be loaded as scikit-learn SVM classifiers.
- The folder code contains three folders:
- manual_labeling - code related exclusively to the process of creating the classifiers, manually labeling accounts and the evaluation process in general.
- hpc_code - code that ran on HPC clusters due to resource limitations when using Jupyter Notebooks.
- topic_modeling_notebooks - contains all Jupyter Notebooks used for analysis and figure creation. Some of the Notebooks were migrated to .py files found in hpc_code due to insufficient resources to run the Notebooks on non-super-computing machines.
- The folder graphics contains all figures shown in the paper and in the appendix.
Bibtex:
@article{elyashar2021state,
title={The State of Mind of Healthcare Professionals in the Light of the COVID-19: Insights from Text Analysis of Twitter's Online Discourses.},
author={Elyashar, Aviad and Plochotnikov, Ilia and Cohen, Idan-Chaim and Puzis, Rami and Cohen, Odeya},
journal={Journal of Medical Internet Research},
year={2021}
}
PMID: 34550899 DOI: 10.2196/30217 Link: 10.2196/30217