We use HLMs to measure urban/rural differences in stress expressions in China after text-mining from Weibo posts that discuss stress. HLMs allow us to model the contribution of regional characteristics on individuals language.
The description of the data and model is part of our paper published at AAAI ICWSM 2022.
Preprint: (TBD)
Supplementary_Materials.pdf: This file contains the supplementary materials for our paper./county_level_main: This directory contains data and code to run the main county-level HLM./county_level_post_stratified: This directory contains the data and code to run the HLM on the county-level with post-stratified weights on gender./province_level: This directory contains data and code to run HLM on the province-level/reference_data: This directory contains files to view reference data, such as the words in each LDA topic, LIWC categories, frequency occurrenecs of phrases, and province gender ratios used to calculate post-stratified weights.
Each directory with HLM scripts should have a corresponding bash file to run to retrieve correlations from the HLM model. These results are also already included in the repo in the /script_results folder for each directory that includes a script. Raw data is in the /data folder and the HLM code itself is in a corresponding hlm_model.R file.
Run the bash script as follows to run HLM:
bash run_hlm.bash
@article{cui2022icwsm,
title={Social Media Reveals Urban-Rural Differences in Stress across China},
author={Cui, Jesse and Zhang, Tingdan and Pang, Dandan and Jaidka, Kokil and Sherman, Garrick and Jakhetiya, Vinit and Ungar, Lyle, and Guntuku, Sharath Chandra},
journal={Proceedings of the International AAAI Conference on Web and Social Media},
year={2022}
}
APA
Cui, J., Zhang, T., Pang, D., Jaidka, D., Sherman, G., Jakhetiya, V., Ungar, L., & Guntuku, S. C. (2022). Social Media Reveals Urban-Rural Differences in Stress across China. Proceedings of the International AAAI Conference on Web and Social Media (ICWSM)