├── utils.R
This file loads the packages and the commonly used functions. Run this script before any other script you want to evaluate.
├── data-upload.R
├── data-setup.R
We loaded the enclosed data and exported that into a private OneDrive folder via the awesome {pins} package. We also saved all intermediate results in this folder. Please e-mail us if you wish to access the folder: granat.marcell@uni-neumann.hu
├── 01-extract_nonenglish.R
├── 02-translate.py
├── 03-merge_translation.R
As a first, we extracted the non-english sentences from the job posts and trasnlated them via Python.
├── 04-sbert.R
├── 04-sbert.py
├── 05-bert_matching.R
We used Bert model to sort the jobs into ISCO categories...
├── 07-bert-skills.R
├── 07-best-skills.py
... and to identify the skills
├── 08-distance.R
Calculates the distance among the regions of Saudi Arabia (see sau_adm_gadm_20210525_shp/).
├── 09-skill-order.R
Calculate the number of incidence of specific skills in the documents (07-bert-skills.R)
├── 10-job-requirements.R
We applied several RegEx filters to detect whether a given job has specific requirements.
├── 11-recommendation.R
├── 12-filter-recommendation.R
Our final merge algorithm.
├── dashboard.rmd
A flexdashboard to visualize our results.
├── sau_adm_gadm_20210525_shp
├── ...
Shape files for Saudi Arabia.