To rerun the study:
Create a virtual environment and install the requirements:
python3.10 -m venv .venv
source .venv/bin/activate
python -m pip install -U pip setuptools wheel
python -m pip install -r requirements.txtDownload the corpus and unpack it into the folder ./data/raw.
Run the Python scripts:
1_parse_raw_data.py - parse the corpus into text files which will be save in ./data/parsed.
2_lemmatize_parsed_data.py - extract lemmas from the text files and save them into ./data/lemmatized as JSON.
3_compute_counts_on_lemmas - count the number of matches for each element of MFD-PL and time period. The results are saved in ./data/couunts.
4_analysis.ipynb - a Jupyter notebook to reproduce the analysis.
The folder ./data/results contains the CSV files and timeseries plots.