How michou001/intelligent-recommendation-plataform works
The Intelligent Recommendation Platform serves personalized Top-K recommendations and records user interactions. Its main request path is user → REST API → recommendation controller → feature lookup → selected strategy → catalog enrichment → response; the controller also publishes an interaction event for asynchronous history storage. Separately, a four-stage offline pipeline prepares features for inference. The graph preserves the README’s strategy, decorator, persistence, and messaging capabilities. Some relationships are documented rather than verified from complete source excerpts; those edges cite the README, and unsampled details are not inferred.
Main entry points: recommendation_controller.py Recommendation Controller · api.py REST API · model_factory.py Model Factory · feature_store_manager.py Feature Store Reader
Components 26
Presentation
- Web Interface
src/presentation/static/index.html - Request Schemas
src/presentation/schemas.py
Application
- REST API
src/application/api.py - Recommendation Controller
src/application/recommendation_controller.py - Interaction Ingestion
src/application/interaction_ingestion_service.py - Platform Settings
src/config/settings.py
Recommendation Engine
- Model Factory
src/ai/model_factory.py - Collaborative Filtering
src/ai/strategies/collaborative_filtering_strategy.py - Content-Based Strategy
src/ai/strategies/content_based_strategy.py - Deep Learning Strategy
src/ai/strategies/deep_learning_strategy.py - Popularity Fallback
src/ai/strategies/popularity_strategy.py - Caching Decorator
src/ai/decorators/caching_decorator.py - Logging Decorator
src/ai/decorators/logging_decorator.py - Timing Decorator
src/ai/decorators/timing_decorator.py
Data Preparation
- Feature Store Reader
src/data/feature_store_manager.py - Dataset Ingestion
src/data/pipeline/ingest.py - Data Cleaning
src/data/pipeline/clean.py - Feature Transformation
src/data/pipeline/transform.py - Feature Materialization
src/data/pipeline/materialize.py
Persistence and Messaging
- User Profiles
src/persistence/sql_user_profile_repository.py - Item Catalog
src/persistence/sql_catalog_repository.py - Interaction Event Queue
src/messaging/interaction_event_queue.py - Feedback History Worker
src/messaging/feedback_history_service.py - Interaction History
src/persistence/nosql_interaction_history_repository.py - Embedding Repository
src/persistence/vector_embedding_repository.py
Other components
- User
Connections 30
- User → Web Interface: uses
README.md - Web Interface → REST API: calls endpoints
README.md - REST API → Request Schemas: validates with
README.md - REST API → Recommendation Controller: routes requests
README.md - REST API → Interaction Ingestion: routes interactions
src/application/api.py - REST API → Model Factory: creates strategy
src/application/api.py - REST API → Platform Settings: configures
src/application/api.py - Recommendation Controller → Feature Store Reader: loads features
src/application/recommendation_controller.py - Recommendation Controller → Collaborative Filtering: predicts via
README.md - Recommendation Controller → Content-Based Strategy: predicts via
README.md - Recommendation Controller → Deep Learning Strategy: predicts via
README.md - Recommendation Controller → Popularity Fallback: falls back to
src/application/recommendation_controller.py - Model Factory → Collaborative Filtering: loads registered
src/ai/model_factory.py - Model Factory → Content-Based Strategy: loads registered
src/ai/model_factory.py - Model Factory → Deep Learning Strategy: loads registered
src/ai/model_factory.py - Model Factory → Popularity Fallback: loads registered
src/ai/model_factory.py - REST API → Caching Decorator: builds decorators
README.md - REST API → Logging Decorator: builds decorators
README.md - REST API → Timing Decorator: builds decorators
README.md - Recommendation Controller → User Profiles: reads profile
README.md - Recommendation Controller → Item Catalog: enriches ranking
src/application/recommendation_controller.py - Recommendation Controller → Interaction Event Queue: publishes event
src/application/recommendation_controller.py - Interaction Ingestion → Interaction Event Queue: publishes interaction
src/application/interaction_ingestion_service.py - Interaction Event Queue → Feedback History Worker: delivers events
src/messaging/interaction_event_queue.py - Feedback History Worker → Interaction History: writes history
src/messaging/feedback_history_service.py - Dataset Ingestion → Data Cleaning: writes raw data
README.md - Data Cleaning → Feature Transformation: writes cleaned data
README.md - Feature Transformation → Feature Materialization: writes features
README.md - Feature Materialization → Feature Store Reader: materializes for
README.md - Feature Store Reader → Recommendation Controller: returns features
src/application/recommendation_controller.py