# vishwanath-06/fusionphishguard architecture

> Architecture diagram of the GitHub repository vishwanath-06/fusionphishguard, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/vishwanath-06/fusionphishguard
- Repository: https://github.com/vishwanath-06/fusionphishguard

## Overview

FusionPhishGuard is a research framework for classifying URLs and mobile-web links as legitimate or phishing. Its documented workflow tokenizes inputs, extracts lexical, subword, contextual, and LLM representations, fuses and refines features, models URL sequences, and classifies the result. The README documents these stages and branches, but the listed implementation paths under src/ are absent from the supplied tree; component paths are therefore left null except for the two present notebooks. Connections reflect the documented pipeline, not inspected runtime calls.

Main entry points: [`notebooks/phishing-with-word-embedding-techniques.ipynb`](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-word-embedding-techniques.ipynb) (Word2Vec), [`notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb`](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb) (MiniLM)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_representations["URL Representations"]
  node_tokenizer["URL Tokenizer"]
  node_word2vec["Word2Vec"]
  node_fasttext["FastText"]
  node_bert["BERT"]
  node_roberta["RoBERTa"]
  node_minilm["MiniLM"]
  node_qwen["Qwen"]
  node_falcon["Falcon"]
end

subgraph group_fusion["Feature Fusion"]
  node_attention["Gated Attention"]
  node_se["SE Refinement"]
end

subgraph group_classification["Classification"]
  node_bilstm["BiLSTM"]
  node_classifier["Classifier"]
  node_prediction["Prediction"]
end

node_analyst(("URL Submitter"))

node_analyst -->|"submits URL"| node_tokenizer
node_tokenizer -->|"tokenizes for"| node_word2vec
node_tokenizer -->|"tokenizes for"| node_fasttext
node_tokenizer -->|"tokenizes for"| node_bert
node_tokenizer -->|"tokenizes for"| node_roberta
node_tokenizer -->|"tokenizes for"| node_minilm
node_tokenizer -->|"tokenizes for"| node_qwen
node_tokenizer -->|"tokenizes for"| node_falcon
node_word2vec -->|"provides features"| node_attention
node_fasttext -->|"provides features"| node_attention
node_bert -->|"provides features"| node_attention
node_roberta -->|"provides features"| node_attention
node_minilm -->|"provides features"| node_attention
node_qwen -->|"provides features"| node_attention
node_falcon -->|"provides features"| node_attention
node_attention -->|"fuses into"| node_se
node_se -->|"refines for"| node_bilstm
node_bilstm -->|"models for"| node_classifier
node_classifier -->|"classifies as"| node_prediction

click node_word2vec "https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-word-embedding-techniques.ipynb"
click node_fasttext "https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-word-embedding-techniques.ipynb"
click node_minilm "https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb"
click node_qwen "https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb"

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_tokenizer,node_word2vec,node_fasttext,node_bert,node_roberta,node_minilm,node_qwen,node_falcon toneBlue
class node_attention,node_se toneAmber
class node_bilstm,node_classifier,node_prediction toneMint
class node_analyst toneIndigo
```

## Components

### URL Representations

- **URL Tokenizer**
- [Word2Vec](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-word-embedding-techniques.ipynb): `notebooks/phishing-with-word-embedding-techniques.ipynb`
- [FastText](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-word-embedding-techniques.ipynb): `notebooks/phishing-with-word-embedding-techniques.ipynb`
- **BERT**
- **RoBERTa**
- [MiniLM](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb): `notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb`
- [Qwen](https://github.com/vishwanath-06/fusionphishguard/blob/main/notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb): `notebooks/phishing-with-combinations-w-t-w-llm-t-llm.ipynb`
- **Falcon**

### Feature Fusion

- **Gated Attention**
- **SE Refinement**

### Classification

- **BiLSTM**
- **Classifier**
- **Prediction**

### Other components

- **URL Submitter**

## Connections

- URL Submitter → URL Tokenizer: submits URL (evidence: `README.md`)
- URL Tokenizer → Word2Vec: tokenizes for (evidence: `README.md`)
- URL Tokenizer → FastText: tokenizes for (evidence: `README.md`)
- URL Tokenizer → BERT: tokenizes for (evidence: `README.md`)
- URL Tokenizer → RoBERTa: tokenizes for (evidence: `README.md`)
- URL Tokenizer → MiniLM: tokenizes for (evidence: `README.md`)
- URL Tokenizer → Qwen: tokenizes for (evidence: `README.md`)
- URL Tokenizer → Falcon: tokenizes for (evidence: `README.md`)
- Word2Vec → Gated Attention: provides features (evidence: `README.md`)
- FastText → Gated Attention: provides features (evidence: `README.md`)
- BERT → Gated Attention: provides features (evidence: `README.md`)
- RoBERTa → Gated Attention: provides features (evidence: `README.md`)
- MiniLM → Gated Attention: provides features (evidence: `README.md`)
- Qwen → Gated Attention: provides features (evidence: `README.md`)
- Falcon → Gated Attention: provides features (evidence: `README.md`)
- Gated Attention → SE Refinement: fuses into (evidence: `README.md`)
- SE Refinement → BiLSTM: refines for (evidence: `README.md`)
- BiLSTM → Classifier: models for (evidence: `README.md`)
- Classifier → Prediction: classifies as (evidence: `README.md`)
