Upload README.md

README.md

@@ -1,3 +1,14 @@
+---
+language: en
+tags:
+  - cybersecurity
+  - malicious-url-detection
+  - bert
+  - transformers
+  - phishing-detection
+license: apache-2.0
+---
+
 # Malicious URL Detection Model
 
 > A fine-tuned **BERT-LoRA** model for detecting malicious URLs, including phishing, malware, and defacement threats.
@@ -8,10 +19,10 @@ This model is a **fine-tuned BERT-based classifier** designed to detect **malici
 
 The model classifies URLs into **four categories**:
 
-- ✅ **Benign**
-- 🔴 **Defacement**
-- ⚠️ **Phishing**
-- 🛑 **Malware**
+- **Benign**
+- **Defacement**
+- **Phishing**
+- **Malware**
 
 It achieves **98% validation accuracy** and an **F1-score of 0.965**, ensuring robust detection capabilities.
 
@@ -21,11 +32,16 @@ It achieves **98% validation accuracy** and an **F1-score of 0.965**, ensuring r
 
 ### Use Cases
 
-✔️ Real-time URL classification for cybersecurity tools✔️ Phishing and malware detection for online safety✔️ Integration into browser extensions for instant threat alerts✔️ Security monitoring for SOC (Security Operations Centers)
+- Real-time URL classification for cybersecurity tools
+- Phishing and malware detection for online safety
+- Integration into browser extensions for instant threat alerts
+- Security monitoring for SOC (Security Operations Centers)
 
 ### Limitations
 
-⚠️ May **misclassify short or obfuscated URLs**⚠️ Performance may degrade with **dynamic domain structures**⚠️ Requires **frequent retraining** to adapt to evolving threats
+- May **misclassify short or obfuscated URLs**
+- Performance may degrade with **dynamic domain structures**
+- Requires **frequent retraining** to adapt to evolving threats
 
 ---
 
@@ -104,17 +120,17 @@ print(f"Prediction: {label_map[prediction]}")
 
 ## Deployment Options
 
-### 1️⃣ Streamlit Web App
+### Streamlit Web App
 
 - Deployed on **Streamlit Cloud, AWS, or Google Cloud**.
 - Provides **real-time URL analysis** with a user-friendly interface.
 
-### 2️⃣ Browser Extension (Planned)
+### Browser Extension (Planned)
 
 - **Real-time scanning** of visited web pages.
 - **Dynamic threat alerts** with confidence scores.
 
-### 3️⃣ API Integration
+### API Integration
 
 - REST API for bulk URL analysis.
 - Supports **Security Operations Centers (SOC)**.
@@ -133,17 +149,17 @@ print(f"Prediction: {label_map[prediction]}")
 
 ### Data Source
 
-Dataset sourced from **Kaggle Malicious URLs Dataset**:📌 [Dataset Link](https://www.kaggle.com/datasets/sid321axn/malicious-urls-dataset)
+Dataset sourced from **Kaggle Malicious URLs Dataset**:  
+📌 [Dataset Link](https://www.kaggle.com/datasets/sid321axn/malicious-urls-dataset)
 
 ### BibTeX Citation
 
 ```
 @article{maliciousurl2025,
-  author    = {Your Name},
+  author    = {r3ddkahili},
   title     = {Fine-Tuned BERT for Malicious URL Detection},
   year      = {2025},
-  journal   = {Cybersecurity AI Research},
-  url       = {https://your-research-paper-link.com}
+  institution = {Western Sydney University}
 }
 ```
 
@@ -151,7 +167,7 @@ Dataset sourced from **Kaggle Malicious URLs Dataset**:📌 [Dataset Link](https
 
 ## Future Work
 
-🚀 **Improvements Planned:**
+**Improvements Planned:**
 
 - **Better phishing URL detection** via adversarial training.
 - **Deploying as a real-time browser extension.**
@@ -159,44 +175,3 @@ Dataset sourced from **Kaggle Malicious URLs Dataset**:📌 [Dataset Link](https
 - **Expanding detection to identify zero-day threats.**
 
 ---
-
-## Uploading to Hugging Face
-
-To upload this model to **Hugging Face**, follow these steps:
-
-```bash
-pip install transformers huggingface_hub
-```
-
-```python
-from huggingface_hub import create_repo
-create_repo("your-huggingface-model-name")
-```
-
-```python
-from transformers import AutoModelForSequenceClassification, AutoTokenizer
-from huggingface_hub import HfApi
-
-model_name = "your-huggingface-model-name"
-model = AutoModelForSequenceClassification.from_pretrained("your-local-model-directory")
-tokenizer = AutoTokenizer.from_pretrained("your-local-model-directory")
-
-# Save & Push Model
-model.save_pretrained(f"{model_name}")
-tokenizer.save_pretrained(f"{model_name}")
-
-api = HfApi()
-api.upload_folder(
-    folder_path=f"{model_name}",
-    repo_id=f"your-huggingface-username/{model_name}",
-    repo_type="model",
-)
-```
-
----
-
-## Conclusion
-
-The **Malicious URL Detection Model** provides **state-of-the-art** accuracy for detecting **phishing, malware, and defacement threats**. It is optimized for **real-time cybersecurity applications** and **deployed using Streamlit**.
-
-✅ **Final F1-score: 0.965**✅ **Optimized for real-time detection**✅ **Ready for deployment via API & browser extension**