This repository contains several notebooks that were used for the project for the course Computational Dialog Modelling. It also contains several figures that were used in the report, and/or referred to from the report.
main.ipynb
includes code to extract the data from the CHILDES corpus,
functions to calculate several turn-based metrics,
and code to generate recurrence matrices and to compute recurrence rates.
Additional measures can be found in main_additional_measures.ipynb
.
Both of these notebooks employ several helper functions located in the tools
folder.
Code for semantic alignment can be found in notebook semantic_recurrence.ipynb
.
Functions used to prepare all corpus data for GLM fits can be found in another notebook, titled
GLM_data.ipynb
.
R code for the linear regressions testing change in complexity over time, and for the GLM testing effects of alignment on complexity is in linear_regressions.r
and glm_models.r
, respectively.
Generated plots can be found in the folder figures
.
- Marianne de Heer Kloots
- Kayleigh Douglas
- Ard Snijders