ChatGPT_Team2 Korean Dialect Understanding of ChatGPT

This is the result of the project "ChatGPT_Team2 Korean Dialect Understanding of ChatGPT"

main folders

aihub dataset main (ver1).py & (ver2).py: use AI Hub dataset to get ChatGPT's response, the result is in test_ver1_90.json and test_ve2_90.json
analysis.py: to analyze generated results, one of the result is analysis.csv
generate_dataset.ipynb: filter the original AI hub dataset, generate useful dataset as result.json
grammar_experiment.py: a few experiments, giving the grammar feature as input
score.py: to score the response
- get the input as naive_result.csv, test.csv
- generate naive_score.csv, score.csv

Related to the JIT dataset

analysis.py: to analyze generated results, one of the result is analysis.csv
change_dataset_format.py: change the existing dataset format similar to the previous AI Hub dataset, generate train_result.json based on train.json
archive_dataset_main.py: to get ChatGPT's response
score.py: to score the response
- get the input as archive_result_vanilla.csv, archive_result.csv,archive_result_with_features.csv
- generate vanilla_score.csv, archive_score.csv, feature_score.csv