Video Recommendation Engine

A FastAPI-based video recommendation system that connects to the EmpowerVerse API to provide personalized video recommendations.

Features

• Personalized Recommendations: Get video recommendations based on user interactions and preferences
• API Caching: Reduces load on the EmpowerVerse API and improves response times
• Fallback System: Gracefully handles API errors by using cached or synthetic data
• Advanced Recommendation Algorithm: Uses content-based and collaborative filtering approaches

Installation

1. Clone the repository:

   git clone https://github.com/Sanaapathann/video-recommendation-assignment/edit/main/README.md
   cd video-recommendation-assignment
   

2. Create and activate a virtual environment:

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

3. Install dependencies:

   pip install -r requirements.txt
   

Usage

Starting the Server

Run the application with:

python run.py

The server will start on the configured port (default: 8000) or find an available port automatically if the default is in use.

API Endpoints

• GET /api/feed?username={username}: Get personalized recommendations for a specific user
• GET /api/kinha-feed: Get recommendations specifically for user "kinha"
• GET /api/universal-feed?username={username}: Get recommendations for any username (real or synthetic)

Control Endpoints

• GET /api/test: Test connection to the EmpowerVerse API
• GET /api/fallback?enable={true|false}: Enable/disable fallback mode
• GET /api/cache?enable={true|false}: Enable/disable caching
• GET /api/cache/clear: Clear all cached data
• GET /api/cache/clear?key={cache_key}: Clear specific cached data

Configuration

The application can be configured through environment variables or the .env file:

Variable	Description	Default
API_BASE_URL	EmpowerVerse API URL	https://api.socialverseapp.com
FLIC_TOKEN	Authentication token	-
PORT	Server port	8000
HOST	Server host	127.0.0.1
REQUEST_TIMEOUT	API request timeout in seconds	5.0
USE_FALLBACK	Enable fallback mode	true
USE_CACHE	Enable caching	true
CACHE_TTL	Cache time-to-live in seconds	3600

Recommendation Algorithm

The recommendation engine uses a combination of:

• User interaction history (views, likes, ratings)
• Content similarity (tags, categories, project codes)
• Creator preferences
• Mood-based matching
• Popularity metrics
• Content recency

Troubleshooting

Port Already in Use

If the default port is already in use, the application will automatically find an available port. You can also specify a different port in the .env file or as an environment variable.

API Connection Issues

If the application cannot connect to the EmpowerVerse API, it will use fallback data or cached responses. You can check the API connection status with the /api/test endpoint.

Data Caching

The application caches API responses to reduce load on the EmpowerVerse API and improve response times. Cached data expires after 1 hour by default.