https://github.com/ycz425/git-net
https://github.com/ycz425/git-net
data-analysis-python graph recommendation
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/ycz425/git-net
- Owner: ycz425
- License: mit
- Created: 2024-12-13T02:56:08.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-12-22T18:02:20.000Z (6 months ago)
- Last Synced: 2024-12-22T18:18:08.361Z (6 months ago)
- Topics: data-analysis-python, graph, recommendation
- Language: Python
- Homepage:
- Size: 9.31 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# GitNet: GitHub Graph Analysis and Recommendation
This project explores the relationships between GitHub users and repositories through graph analysis. Leveraging the GitHub API, it constructs and analyzes a graph where:
- **Nodes** represent repositories and users.
- **Edges** represent relationships such as:
- Forking relationships between repositories.
- Starring relationships between users and repositories.
## Features
### Graph Construction
- **Repositories as Nodes**: Each GitHub repository is represented as a node.
- **Users as Nodes**: Each GitHub user is represented as a node.
- **Edges**:
- Between repositories: Representing forking relationships.
- Between users and repositories: Representing starring relationships.
### Data Analysis
The project includes the following graph analysis features:
- **Connected Components Analysis**: Identifying distinct clusters of nodes within the graph.
- **Degree Centrality** (not available through UI): Measuring the importance of nodes based on the number of direct connections.
- **Betweenness Centrality** (not available through UI): Identifying nodes that act as bridges between different parts of the graph.
- **Louvain Community Detection**: Visualizing communities within the graph using modularity-based clustering.
### Recommendation System
- A hybrid recommendation algorithm:
- **Collaborative Filtering**: Suggesting repositories based on similar users.
- **Graph-Based Recommendations**: Suggesting neighboring repositories based on graph structure.
### Visualization
- Basic visualizations using **Plotly** to:
- Display the graph structure.
- Highlight communities and connected components.
- Further improvements to visualizations are planned.
## Usage
### Prerequisites
- Python 3.7+
- GitHub fine-grained access token (for fetching data from API).
### Installation
1. Clone the repository:
```bash
git clone https://github.com/ycz425/git-net.git
cd git-net
```
2. Install required dependencies:
```bash
pip install -r requirements.txt
```
### Running the Project
1. Create a .env file and set up your GitHub fine-grained access token as an environment variable (only necessary for fetching data from GitHubAPI):
```Python
GITHUB_FINE_GRAINED_ACCESS_TOKEN =
```
2. Run the script to start tbe Dash app:
```bash
python app.py
```
3. Visualize the graph and analysis results in your browser. Click on user nodes to generate recommendations.
4. Explore other algorithms and functions in ```graph.py```.
## Future Plans
- Enhanced visualizations with more interactive features.
- Integration with advanced machine learning models for recommendations.
- Scalability improvements for handling large datasets.
## Contributing
Contributions are welcome! Please feel free to submit issues or pull requests.
## License
This project is licensed under the MIT License. See the `LICENSE` file for more details.
## Acknowledgments
- **NetworkX** for graph analysis.
- **Plotly** for visualizations.
- **Dash** for web interface.
- **GitHub API** for providing access to repository and user data.
---
Feel free to reach out if you have any questions or feedback!