jankounchained / emodynamics-dungeons-and-dragons

NLP Exam 2021

MSc Cognitive Science, Aarhus University

Frida Hæstrup & Jan Kostkan

This repository contains code related to our exam project in the Autumn 2021 course in Natural Language Processing, MSc Cognitive Science, Aarhus University.

Project structure

The repository has the following directory structure:

emodynamics-dungeons-and-dragons/  
├── reproduce_analysis.sh #running analysis
├── checks_windows.sh #running check on window size
├── checks_bins.dh #running check on bin size
├── emissions.csv #estimated carbon emissions from analysis
├── src/  #folder with analysis scripts 
│   └── analysis/ #models
│   │   └── ...
│   └── entropies/ #infodynamics
│   │   └── ...
│   └── binning.py #timebinning
│   └── classification.py #emotion classification
│   └── ntr.py #calculate novelty, transience, resonance
│   └── util.py #utility functions

Environmental impact

The estimated emissions of this analysis can be seen in emissions.csv. Power consumption of this analysis and its experiments was 0.017 kWh. Based on Codecarbon’s estimation of carbon efficiency in Denmark, where the analysis was run, the total emissions are estimated to be 0.01 kgCO_2 eq.