The dataset of the paper titled "Automated Comment Update: How Far are We?".

This is the online repository of the paper "Automated Comment Update: How Far are We?" appeared in ICPC'2021. We release the source code of H_EBC_UP, the dataset used in our evaluation, as well as the experiment results.

• Dataset: The cleaned dataset from Liu's ASE20 paper (i.e., Automating Just-In-Time Comment Updating). We removed the three types of noisy data in the dataset as described in our paper.

• Results:

  • Baseline_CUP.json: The readable result files of CUP.
  • HebCup_all.json: The readable result files of H_EBC_UP.
  • HebCup_correct.json_: The correctly updated items of H_EBC_UP.

• Source: The source code for running H_EBC_UP. 

Run H_EBC_UP:

python3 HebCup.py --dataPath path2dataset

By default, we store the dataset in ./dataset