# dattatejaofficial/url-phishing-detection architecture

> Architecture diagram of the GitHub repository dattatejaofficial/url-phishing-detection, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-19). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/dattatejaofficial/url-phishing-detection
- Repository: https://github.com/dattatejaofficial/url-phishing-detection

## Overview

This is a phishing-URL detection product: browser users submit URLs for real-time classification, while a separate training pipeline builds and publishes models. A monitoring and feedback loop is documented to prompt retraining. The main training-stage calls are visible in the pipeline excerpt; API and service relationships are supported by the source index and excerpts. Some monitoring and integration details are documented only in the README, so those links are marked as documented rather than inferred from unsampled code.

Main entry points: [`phishingsystem/pipeline/training_pipeline/pipeline.py`](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/training_pipeline/pipeline.py) (Training Orchestrator), [`phishingsystem/pipeline/retraining_pipeline/pipeline.py`](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/retraining_pipeline/pipeline.py) (Retraining Pipeline), [`app/services.py`](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/services.py) (Inference Services), [`phishingsystem/pipeline/monitoring_pipeline/monitoring`](https://github.com/dattatejaofficial/url-phishing-detection/tree/main/phishingsystem/pipeline/monitoring_pipeline/monitoring) (Drift Monitoring)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_inference["Live Detection"]
  node_extension["PhishGuard Extension"]
  node_api["REST API<br/>[app.py]"]
  node_inference_services["Inference Services<br/>[services.py]"]
  node_url_cleaner["URL Cleaning<br/>[url_cleaner.py]"]
  node_url_features["URL Features"]
end

subgraph group_training["Model Training"]
  node_training_pipeline["Training Orchestrator<br/>[pipeline.py]"]
  node_data_preparation["Data Preparation"]
  node_feature_extraction["Feature Extraction"]
  node_training_features["URL Feature Logic"]
  node_data_validation["Data Validation<br/>[data_validation.py]"]
  node_data_persistence["Data Persistence"]
  node_data_transformation["Data Transformation"]
  node_model_training["Model Training<br/>[model_training.py]"]
  node_model_evaluation["Model Evaluation"]
  node_model_finalizer["Model Finalization<br/>[model_finalizer.py]"]
  node_artifact_publisher["Artifact Publisher"]
  node_retraining_pipeline["Retraining Pipeline<br/>[pipeline.py]"]
end

subgraph group_feedback["Feedback Monitoring"]
  node_feedback_service["Feedback Service<br/>[services.py]"]
  node_mongo_feedback[("MongoDB Feedback")]
  node_monitoring["Drift Monitoring"]
  node_monitor_utils["Monitoring Utilities"]
  node_retraining_decision["Retraining Decision<br/>[controller.py]"]
end

subgraph group_shared["Shared Model Data"]
  node_model_store[("Production Model")]
  node_azure_storage[("Azure Blob Storage")]
end

node_browser_user(("Browser User"))

node_browser_user -->|"browses with"| node_extension
node_extension -.->|"requests prediction"| node_api
node_api -->|"uses services"| node_inference_services
node_inference_services -->|"cleans URL"| node_url_cleaner
node_inference_services -->|"extracts features"| node_url_features
node_inference_services -->|"loads model"| node_azure_storage
node_inference_services -->|"predicts with"| node_model_store
node_browser_user -.->|"submits feedback"| node_feedback_service
node_feedback_service -->|"stores feedback"| node_mongo_feedback
node_training_pipeline -->|"invokes"| node_data_preparation
node_training_pipeline -->|"invokes"| node_feature_extraction
node_feature_extraction -->|"uses feature logic"| node_training_features
node_training_pipeline -->|"invokes"| node_data_validation
node_training_pipeline -->|"invokes"| node_data_persistence
node_training_pipeline -->|"invokes"| node_data_transformation
node_training_pipeline -->|"invokes"| node_model_training
node_training_pipeline -->|"invokes"| node_model_evaluation
node_training_pipeline -->|"invokes"| node_model_finalizer
node_training_pipeline -->|"invokes"| node_artifact_publisher
node_monitoring -->|"uses utilities"| node_monitor_utils
node_mongo_feedback -.->|"provides recent feedback"| node_monitoring
node_monitoring -->|"evaluates reports"| node_retraining_decision
node_retraining_decision -.->|"dispatches retraining"| node_retraining_pipeline
node_retraining_pipeline -->|"invokes"| node_data_persistence
node_retraining_pipeline -->|"invokes"| node_data_validation
node_retraining_pipeline -->|"invokes"| node_data_transformation
node_retraining_pipeline -->|"invokes"| node_model_training
node_retraining_pipeline -->|"invokes"| node_model_evaluation
node_retraining_pipeline -->|"invokes"| node_model_finalizer
node_retraining_pipeline -->|"invokes"| node_artifact_publisher
node_artifact_publisher -.->|"publishes artifacts"| node_azure_storage

click node_extension "https://github.com/dattatejaofficial/url-phishing-detection/tree/main/extension"
click node_api "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/app.py"
click node_inference_services "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/services.py"
click node_url_cleaner "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/utils/url_cleaner.py"
click node_url_features "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/utils/url_feature_extraction.py"
click node_feedback_service "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/services.py"
click node_training_pipeline "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/training_pipeline/pipeline.py"
click node_data_preparation "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_preparation.py"
click node_feature_extraction "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/feature_extraction.py"
click node_training_features "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/utils/url_preparation/url_feature_extraction.py"
click node_data_validation "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_validation.py"
click node_data_persistence "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_persistance.py"
click node_data_transformation "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_transformation.py"
click node_model_training "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_training.py"
click node_model_evaluation "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_evaluation.py"
click node_model_finalizer "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_finalizer.py"
click node_artifact_publisher "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/artifact_publisher.py"
click node_monitoring "https://github.com/dattatejaofficial/url-phishing-detection/tree/main/phishingsystem/pipeline/monitoring_pipeline/monitoring"
click node_monitor_utils "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/monitoring_pipeline/monitoring_utils.py"
click node_retraining_decision "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/monitoring_pipeline/retraining_controller/controller.py"
click node_retraining_pipeline "https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/retraining_pipeline/pipeline.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_extension,node_api,node_inference_services,node_url_cleaner,node_url_features,node_browser_user toneBlue
class node_training_pipeline,node_data_preparation,node_feature_extraction,node_training_features,node_data_validation,node_data_persistence,node_data_transformation,node_model_training,node_model_evaluation,node_model_finalizer,node_artifact_publisher,node_retraining_pipeline toneAmber
class node_feedback_service,node_mongo_feedback,node_monitoring,node_monitor_utils,node_retraining_decision toneMint
class node_model_store,node_azure_storage toneRose
```

## Components

### Live Detection

- [PhishGuard Extension](https://github.com/dattatejaofficial/url-phishing-detection/tree/main/extension): `extension`
- [REST API](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/app.py): `app/app.py`
- [Inference Services](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/services.py): `app/services.py`
- [URL Cleaning](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/utils/url_cleaner.py): `app/utils/url_cleaner.py`
- [URL Features](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/utils/url_feature_extraction.py): `app/utils/url_feature_extraction.py`

### Model Training

- [Training Orchestrator](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/training_pipeline/pipeline.py): `phishingsystem/pipeline/training_pipeline/pipeline.py`
- [Data Preparation](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_preparation.py): `phishingsystem/components/data_preparation.py`
- [Feature Extraction](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/feature_extraction.py): `phishingsystem/components/feature_extraction.py`
- [URL Feature Logic](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/utils/url_preparation/url_feature_extraction.py): `phishingsystem/utils/url_preparation/url_feature_extraction.py`
- [Data Validation](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_validation.py): `phishingsystem/components/data_validation.py`
- [Data Persistence](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_persistance.py): `phishingsystem/components/data_persistance.py`
- [Data Transformation](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/data_transformation.py): `phishingsystem/components/data_transformation.py`
- [Model Training](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_training.py): `phishingsystem/components/model_training.py`
- [Model Evaluation](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_evaluation.py): `phishingsystem/components/model_evaluation.py`
- [Model Finalization](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/model_finalizer.py): `phishingsystem/components/model_finalizer.py`
- [Artifact Publisher](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/components/artifact_publisher.py): `phishingsystem/components/artifact_publisher.py`
- [Retraining Pipeline](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/retraining_pipeline/pipeline.py): `phishingsystem/pipeline/retraining_pipeline/pipeline.py`

### Feedback Monitoring

- [Feedback Service](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/app/services.py): `app/services.py`
- **MongoDB Feedback**
- [Drift Monitoring](https://github.com/dattatejaofficial/url-phishing-detection/tree/main/phishingsystem/pipeline/monitoring_pipeline/monitoring): `phishingsystem/pipeline/monitoring_pipeline/monitoring`
- [Monitoring Utilities](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/monitoring_pipeline/monitoring_utils.py): `phishingsystem/pipeline/monitoring_pipeline/monitoring_utils.py`
- [Retraining Decision](https://github.com/dattatejaofficial/url-phishing-detection/blob/main/phishingsystem/pipeline/monitoring_pipeline/retraining_controller/controller.py): `phishingsystem/pipeline/monitoring_pipeline/retraining_controller/controller.py`

### Shared Model Data

- **Production Model**
- **Azure Blob Storage**

### Other components

- **Browser User**

## Connections

- Browser User → PhishGuard Extension: browses with
- PhishGuard Extension → REST API: requests prediction
- REST API → Inference Services: uses services (evidence: `app/app.py`)
- Inference Services → URL Cleaning: cleans URL (evidence: `app/services.py`)
- Inference Services → URL Features: extracts features (evidence: `app/services.py`)
- Inference Services → Azure Blob Storage: loads model (evidence: `app/services.py`)
- Inference Services → Production Model: predicts with (evidence: `app/services.py`)
- Browser User → Feedback Service: submits feedback
- Feedback Service → MongoDB Feedback: stores feedback (evidence: `app/services.py`)
- Training Orchestrator → Data Preparation: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Feature Extraction: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Feature Extraction → URL Feature Logic: uses feature logic (evidence: `phishingsystem/components/feature_extraction.py`)
- Training Orchestrator → Data Validation: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Data Persistence: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Data Transformation: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Model Training: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Model Evaluation: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Model Finalization: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Training Orchestrator → Artifact Publisher: invokes (evidence: `phishingsystem/pipeline/training_pipeline/pipeline.py`)
- Drift Monitoring → Monitoring Utilities: uses utilities (evidence: `phishingsystem/pipeline/monitoring_pipeline/monitoring/__init__.py`)
- MongoDB Feedback → Drift Monitoring: provides recent feedback
- Drift Monitoring → Retraining Decision: evaluates reports
- Retraining Decision → Retraining Pipeline: dispatches retraining (evidence: `phishingsystem/pipeline/monitoring_pipeline/retraining_controller/controller.py`)
- Retraining Pipeline → Data Persistence: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Data Validation: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Data Transformation: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Model Training: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Model Evaluation: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Model Finalization: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Retraining Pipeline → Artifact Publisher: invokes (evidence: `phishingsystem/pipeline/retraining_pipeline/pipeline.py`)
- Artifact Publisher → Azure Blob Storage: publishes artifacts
