PySocialLab

PySocialLab is a pure Python project for cleaning, exploring, and recommending in social networks. It uses only the Python standard library and Jupyter notebooks to process JSON data describing users, friendships, and liked pages.

Features

• Clean raw social data (remove duplicates, inactive users/pages)
• Recommend pages based on common user interests
• Recommend people via mutual friend connections
• 100% pure Python — no external dependencies

Notebooks

• data_cleaning.ipynb: Load and clean raw data, output cleaned JSON
• data_visualization.ipynb: Display users, connections, and pages from loaded data
• pages_yml.ipynb: Recommend pages a user might like based on similar users' interests
• people_ymk.ipynb: Recommend people ("People You May Know") via mutual friend connections

Getting Started

Prerequisites: Python 3.7+ and Jupyter (jupyterlab or notebook).

python3 -m venv .venv
source .venv/bin/activate
pip install jupyter
jupyter lab

Usage

1. Run data_cleaning.ipynb to clean raw JSON data and output cleaned_data2.json.
2. Run data_visualization.ipynb to view users, connections, and pages.
3. Use pages_yml.ipynb to get page recommendations for a user ID.
4. Use people_ymk.ipynb to get friend recommendations for a user ID.

All work is done in pure Python within notebooks.