This project looks into calculating stylistic and lexical alignment in multi-party multi-gender dialogues using generalised linear models.

Make sure to activate the conda environment cdm (or first create one with environment.yml).

To calculate stylistic alignment, run stylistic.py with the arguments for dataset and the type of experiment to perform. For example,

python stylistic.py AMI 1

calculates performs Experiment 1 for stylistic alignment for the AMI dataset.

It is also possible to perform Experiment 1 for stylistic alignment per linguistic category. To do so, set the --cat argument to True.

To calculate lexical alignment, run lexica.py with the arguments for dataset and type of experiment, as before.