AdvOC

This repository is an implementation of paper: Data Quality Matters: A Case Study of Obsolete Comment Detection

Requirements

• openjdk11 for tree data generation
• pip install -r requirements.txt

Data

The original dataset is downloaded from here

The cleaned dataset provided by us can be downloaded from here

All scripts used for data processing are already in the data_process directory, you can use them to process original dataset or generate new datasets yourself.

cd data_process
./original_to_cleaned.sh

Train a new model

First, modify the DATA_PATH in constants.py. Then run

CUDA_VISIBLE_DEVICES=0 python3 run_AdvOC.py --seq --tree --script --model_path=model

Evaluate the model

CUDA_VISIBLE_DEVICES=0 python3 run_AdvOC.py --seq --tree --script --model_path=model --test_mode

We provide a trained model classifier.pt, together with the dataset. you can run it without training

CUDA_VISIBLE_DEVICES=0 python3 run_AdvOC.py --seq --tree --script --model_path=model --model_file=/path/to/data/classifier.pt --test_mode