Pixly: AI-assisted Movie Recommendation Engine

Pixly is a sophisticated movie recommendation platform that leverages advanced collaborative filtering and machine learning techniques using MovieLens data. Originally developed as a learning project, it evolved into a full-featured content platform with rich social features and advanced recommendation capabilities.

The basic idea relies on collaborative filtering for film recommendations. TMDB and MovieLens data are used for user-similarity computations. Roughly 40.000 people who had voted for more than 20 movies in Movielens were randomly selected. When a new user votes for 20 films, people from the database with similar cinema tastes (highly correlated people) are found. These highly similar people were assigned to this new user in less than a minute. Afterward, a score prediction for a movie not seen yet (on average +/- 0.5 points) is computed for the user when a new film page is visited. Also, various dimensionality reduction techniques are researched and used for clustering the movies.

Core Features

Recommendation Engine

• Collaborative filtering based on user similarity computations
• Content-based filtering using TMDB and MovieLens data
• Real-time prediction engine (accuracy: +/- 0.5 points)
• Sophisticated user similarity calculations for personalized recommendations
• Multi-threaded computation optimization using Cython
• Advanced clustering using dimensionality reduction techniques

Content Management

• Comprehensive movie metadata management
• Rich text content support with CKEditor integration
• Multiple image format handling (posters, covers, hero images)
• Tag-based categorization system
• SEO-optimized content structure
• Blog system with rich media support

Social Features

• User profiles and personalization
• Custom lists and collections
• Rating and review system
• Social media integration
• User interaction capabilities

Search and Discovery

• Elasticsearch-powered search
• Advanced filtering options
• Tag-based browsing
• SEO-optimized content pages

Technical Architecture

Backend (Django)

• Core Framework: Django with GraphQL API
• Performance Optimizations:
  • Cython-optimized computational modules
  • Redis caching layer
  • Multi-threaded processing
  • Efficient database queries
• Data Processing:
  • NumPy for numerical computations
  • Custom algorithm implementations
  • MovieLens data integration
  • TMDB API integration

Frontend

• Main Application: React-based SPA
• Blog Platform: Svelte implementation
• Design System: Atomic design with styled-components
• API Integration: Apollo GraphQL client

Cloud Infrastructure (AWS)

• Hosting: Elastic Beanstalk
• Database:
  • RDS (MySQL)
  • Elasticache (Redis)
• Storage: S3 for media files
• Security:
  • SSL/TLS encryption
  • CORS configuration
  • Secure cookie handling

External Integrations

• TMDB API for movie data
• YouTube API for video content
• Social media platforms
• Search engine optimization tools

Tech Stack

Cloud Services

• AWS Elastic Beanstalk
• AWS Elasticache
• AWS RDS
• AWS S3

Backend Technologies

• Python Django
• Redis
• MySQL
• Graphene (GraphQL)
• Cython
• Numpy
• Elasticsearch

Frontend Technologies

• React
• Apollo GraphQL
• Styled Components
• CKEditor

Project Structure

pixly-app/
├── algorithm/      # Recommendation engine & Cython optimizations
├── djaws/         # Core Django configuration
├── gql/           # GraphQL API implementation
├── items/         # Content management models
├── persona/       # User recommendation system
├── persons/       # User management & profiles
├── blog/          # Blog functionality
├── rss/           # RSS feed functionality
└── _FRONTEND/     # React application

---

Contact Details: [email protected]

Python Modules Documentation

Core Modules

manage.py

• Django's command-line utility for administrative tasks
• Configures Django environment with 'djaws.settings'
• Entry point for running development server and migrations

djaws/

• settings.py: Core Django configuration with AWS integration
• urls.py: Main URL routing configuration
• wsgi.py: WSGI application entry point
• cognito.py: AWS Cognito integration for authentication
• storage_backends.py: AWS S3 storage configuration
• common_views.py: Shared view functionality

Feature Modules

algorithm/

• models.py: Database models for algorithm data
• custom_functions.py: Custom algorithmic implementations
• statistics.py: Statistical analysis functions
• movielens.py: MovieLens dataset integration
• cython/: Performance-optimized Cython implementations
  • calculations.pyx: Core computational functions
  • setup.py: Cython build configuration

persona/

• models.py: User persona and preference models
• myqueue.py: Queue management for recommendations
• profile.py: User profile functionality
• views.py: Persona-related view logic

persons/

• models.py: User and authentication models
• profile.py: Extended profile functionality
• abstract.py: Abstract base classes
• views.py: User management views

items/

• models.py: Content and media models
• tmdb.py: TMDB API integration
• forms.py: Form definitions
• views.py: Content management views
• outerApi.py: External API integrations

gql/

• schema.py: GraphQL schema definition
• types.py: GraphQL type definitions
• mutations.py: GraphQL mutations
• search.py: Search functionality
• tmdb_class.py: TMDB data integration
• elasticsearch_client.py: Elasticsearch integration
• social/: Social media integrations
  • social.py: Base social functionality
  • facebook.py: Facebook integration
  • twitter.py: Twitter integration

blog/

• models.py: Blog content models
• views.py: Blog rendering views
• urls.py: Blog URL routing
• sitemaps.py: SEO sitemap generation

rss/

• models.py: RSS feed models
• views.py: Feed generation views

Utility Modules

pixly/

• lib.py: Common utility functions
• cache_functions.py: Caching implementations
• indexing.py: Search indexing functionality
• google_index_search.py: Google search integration