Github repository for EMNLP 2020 paper Condolence and Empathy in Online Communities by Naitian Zhou and David Jurgens.
Due to the sensitive nature of this data, it will be made available upon request to researchers.
If you would like to use our pretrained models, you can install with pip with
pip install condolence-models
See the Github repository or the PyPi listing for more details and documentation.
Notebooks for analysis on the final Reddit 2018 dataset are in the analysis
folder. Notebooks are named by the order in which they are to be run. You can
also view notebook dependencies at the top of each notebook.
Some requisite auxiliary data has also been included (namely, Reddit 2018 daily volume).
Code for crawling through Reddit to heuristically label condolence and distress
are located in the crawling directory. The directory also includes the
labeling script used to get the final Reddit 2018 dataset used for analysis
(grab_comments.py).
Note: this code might be messy. Peruse at your own risk.
Notebooks and code for training models are in the models directory. This code
might be messy. Peruse at your own risk.
Naitian Zhou (naitian@umich.edu)