PAPEA-Pipeline

This is the development repository for scripts based on the PAPEA paper. It contains current versions of the scripts used in the original PAPEA paper and additional scripts.

The Python and R scripts are located in the folders 'python_scripts' and 'r_scripts'

• 'python_scripts/1_papea_pipeline_python.ipynb' is a Jupyter notebook for the first part of the PAPEA pipeline
• 'python_scripts/Actor Identification.ipynb' is a Jupyter notebook actor identification in protest form sentences using LLMs
• 'python_scripts/counterprotest_prediction.ipynb' is a Jupyter notebook for prdicting whether a protest event is a counter protest
• 'r_scripts/2_papea_pipeline_R.Rmd' is a R markdown script for the second part of the PAPEA pipeline
• 'r_scripts/3_papea_evaluate_predictions.Rmd' is a R markdown script replicating the evaulation of the performance of the language models in Table 1b and Table 2
• 'r_scripts/4_papea_appendix.Rmd' is a R markdown script to replicate all computations in the appendix.