Social Platform

Social is a modern social media platform designed to provide personalized content recommendations and seamless user interactions.

Features

• Personalized Recommendations: AI-powered recommendations for timeline, explore, and follow suggestions.
• Post View Tracking: Tracks posts users have seen to avoid showing duplicate content.
• Infinite Scroll: Smooth infinite scrolling for timeline and explore pages.
• User Profiles: View and manage user profiles with posts and follower details.
• Content Categorization: Posts categorized by user interests for better discovery.
• Authentication: Secure login and user management.
• Dockerized Environment: Simplifies development and deployment with Docker.
• TypeScript Support: Ensures type safety and modern JavaScript features.

Project Structure

├── app.ts
├── dev.sh
├── docker-compose.override.yml
├── docker-compose.yml
├── Dockerfile
├── Dockerfile.dev
├── package.json
├── pnpm-lock.yaml
├── prod.sh
├── README.md
├── tsconfig.json
├── algorithm/
│   ├── app/
│   │   ├── main.py
│   │   ├── model.py
│   │   ├── utils.py
├── app/
│   ├── controllers/
│   ├── lib/
│   ├── middleware/
│   ├── models/
│   ├── public/
│   ├── router/
│   ├── services/
│   ├── scripts/

Getting Started

Prerequisites

• Node.js
• Docker
• Python 3.11+

Installation

1. Clone the repository:

   git clone https://github.com/your-repo/social-recommender.git

2. Install dependencies:

   pnpm install

3. Set up environment variables:

   • Create a .env file in the root directory.
   • Add necessary configurations for MongoDB, Python service, and other dependencies.

4. Start the services:

   docker-compose up --build

Key Endpoints

Node.js API

• /timeline: Fetches the user's timeline with personalized recommendations.
• /ai/explore: Provides AI-powered explore page recommendations.
• /ai/users/recommended: Suggests users to follow based on interests and connections.
• /user/seen-posts: Tracks and manages posts seen by the user.

Python Service

• /recommend/timeline/{user_id}: AI recommendations for the timeline.
• /recommend/explore/{user_id}: AI recommendations for the explore page.
• /recommend/users/{user_id}: AI-powered follow suggestions.

Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository.
2. Create a new branch for your feature or bug fix.
3. Submit a pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.