Video Recommendation Engine

A sophisticated recommendation system that suggests personalized video content based on user preferences and engagement patterns using deep neural networks. Ref: to see what kind of motivational content you have to recommend, take reference from our Empowerverse App ANDROID || iOS.

🎯 Project Overview

This project implements a video recommendation algorithm that:

• Delivers personalized content recommendations
• Handles cold start problems using mood-based recommendations
• Utilizes Graph/Deep neural networks for content analysis
• Integrates with external APIs for data collection
• Implements efficient data caching and pagination

🛠️ Technology Stack

• Backend Framework: FastAPI
• Documentation: Swagger/OpenAPI

📋 Prerequisites

• Virtual environment (recommended)

🚀 Getting Started

1. Clone the Repository

   git clone https://github.com/Tim-Alpha/video-recommendation-assignment.git

   cd video-recommendation-engine

2. Set Up Virtual Environment

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

   pip install -r requirements.txt

4. Configure Environment Variables Create a .env file in the root directory:

   FLIC_TOKEN=your_flic_token
   API_BASE_URL=https://api.socialverseapp.com

5. Run Database Migrations

   alembic upgrade head

6. Start the Server

   uvicorn app.main:app --reload

📊 API Endpoints

Recommendation Endpoints Has to Build

1. Get Personalized Feed

   GET /feed?username={username}
   

   Returns personalized video recommendations for a specific user.

2. Get Category-based Feed

   GET /feed?username={username}&project_code={project_code}
   

   Returns category-specific video recommendations for a user.

Data Collection Endpoints (Internal Use)

APIs for data collection:

APIs

1. Get All Viewed Posts (METHOD: GET):

   https://api.socialverseapp.com/posts/view?page=1&page_size=1000&resonance_algorithm=resonance_algorithm_cjsvervb7dbhss8bdrj89s44jfjdbsjd0xnjkbvuire8zcjwerui3njfbvsujc5if
   

2. Get All Liked Posts (METHOD: GET):

   https://api.socialverseapp.com/posts/like?page=1&page_size=1000&resonance_algorithm=resonance_algorithm_cjsvervb7dbhss8bdrj89s44jfjdbsjd0xnjkbvuire8zcjwerui3njfbvsujc5if
   

3. Get All Inspired posts (METHOD: GET):

   https://api.socialverseapp.com/posts/inspire?page=1&page_size=1000&resonance_algorithm=resonance_algorithm_cjsvervb7dbhss8bdrj89s44jfjdbsjd0xnjkbvuire8zcjwerui3njfbvsujc5if
   

4. Get All Rated posts (METHOD: GET):

   https://api.socialverseapp.com/posts/rating?page=1&page_size=1000&resonance_algorithm=resonance_algorithm_cjsvervb7dbhss8bdrj89s44jfjdbsjd0xnjkbvuire8zcjwerui3njfbvsujc5if
   

5. Get All Posts (Header required*) (METHOD: GET):

   https://api.socialverseapp.com/posts/summary/get?page=1&page_size=1000
   

6. Get All Users (Header required*) (METHOD: GET):

   https://api.socialverseapp.com/users/get_all?page=1&page_size=1000
   

Authorization

For autherization pass Flic-Token as header in the API request:

Header:

"Flic-Token": "flic_e657dc77ce71262a4f5dc78e0c9e1bf02a7f0e186b664708c39369f182ad2518"

Note: All external API calls require the Flic-Token header:

📝 Submission Requirements

1. GitHub Repository

   • Submit a merge request from your fork or cloned repository.
   • Include a complete Postman collection demonstrating your API endpoints.
   • Add a docs folder explaining how your recommendation system works.

2. Video Submission

   • Introduction Video (30–40 seconds)
     • A short personal introduction (with face-cam).
   • Technical Demo (3–5 minutes)
     • Live demonstration of the APIs using Postman.
     • Brief overview of the project. Video Submission

3. Notification

   • Join the Telegram group: Video Recommendation
   • Notify upon completion

✅ Evaluation Checklist

• All APIs are functional
• Database migrations work correctly
• README is complete and clear
• Postman collection is included
• Videos are submitted
• Code is well-documented
• Implementation handles edge cases
• Proper error handling is implemented