FKG Framework

This repository contains the implementation and resources for the FKG Framework, designed to process and analyze datasets effectively. The repository includes preprocessing scripts, code for model training, and datasets for evaluation.

Repository Structure

• FKG: Main module for the FKG framework.
• Preprocess_FKG: Raw data to prepare it for the FKG pipeline.
• code: Python scripts for implementing the core functionalities of the FKG framework, including training and evaluation.
• data: Directory containing the proposed test set (Part_1) and supporting data files for evaluation.
• data_1: Additional dataset directory containing (Part_2) test data for the framework.
• README.md: Documentation file for the repository.

Setup and Usage

Prerequisites

• Python 3.7 or higher
• Required Python libraries:
  • numpy
  • pandas
  • torch
  • tqdm
  • scikit-learn

pip install -r requirements.txt

### Dataset
- The proposed test sets are available in the `data` and `data_1` folders.

### Running the Code
1. Preprocess the data using the scripts in `Preprocess_FKG`:
   ```bash
   python Preprocess_FKG/preprocess.py

2. Train or evaluate models using the scripts in the code folder:

   python code/train_model.py

3. Use the test sets from data and data_1 to validate your results.

Results and Evaluation

• Outputs include preprocessed data, model predictions, and performance metrics.
• Evaluation results are saved in the specified output directory.