Gavel-Replication-Package

Description

This repository hosts the implementation and datasets for two main components: GAVEL and CNBN. CNBN serves as the baseline for our comparisons, while GAVEL is our newly proposed system designed to improve upon the baseline in specific tasks.

Repository Structure

• CNBN (Baseline):

  • gh_action_dataset: Includes datasets like all.csv, code_block.csv, and results from multiple 5-fold cross-validation setups.
  • tools: Contains scripts such as dataLoader.py, main.py, and predictor.py to facilitate data handling and predictions.

• GAVEL (Proposed System):

  • CONFS: Configuration files (C1.csv to C5.csv) for setting up experimental conditions.
  • RESULTS: Contains experimental outputs like Fold_1_Results.csv, utilities for converting classification reports to LaTeX (classification_report2latex.py), and statistical analysis tools (stats.py).
  • TOOLS: Includes scripts like gavel.py which are central to the GAVEL system.