This is the repository of the replication package for paper "Understanding Practitioners’ Expectations on Clear Code Review Comments".
Code/contains the code of our studyData/contains our manually labelled data, the five-fold data for experiments, and extra datasetOthers/contains the sample survey and the paper list we collected
- Model Set 1:
lstm_classifier.pyandml_classifier.py - Model Set 2:
run_evaluator.pywithrun.sh - Model Set 3:
llm.pyandllm_compute_metrics.py augmentor.py: the data augmentation code
crc_labelled_data.csv: the labelled dataset for our main experiments, processed from the CodeReviewer datasetcodereviewer_new_test.csv: the manually labelled subset of the CodeReviewer_New datasetfive_fold_data: the up sampling data for five-fold cross validation
Survey Sample.pdf: the sample survey we used in our studyCode Review Paper List: 47 papers we collected related to CRCs in our study