Final project for info259
-
data_preprocessing: separate notebook for HC3 import and transformations to get dataframe into form: q_id / question / answer / true_label (real/fake) / source
- TO DO: bring "!\nnetwork error" code over from research pipeline notebook
-
research_pipeline: import dataframe from preprocessing step, run detector model and extract / analyze misclassified texts
- TO DO: update Extract Misclassified Data section to run on cleaned dataframe from data_preprocessing notebook
-
shap_analysis: runs detector model on subset of data to obtain SHAP score for quantitative evaluation
-
qual_analysis: