CodeBERT-base-Malicious_URLs Open Source Model - Free Detection of Multi-Class Malicious URLs

Home

Codebert Base Malicious URLs

Developed by DunnBC22

A malicious URL detection model fine-tuned based on microsoft/codebert-base for multi-classification tasks

Text Classification

Transformers

English#Malicious URL Classification #CodeBERT Fine-tuning #Cybersecurity Detection

Downloads 1,308

Release Time : 5/20/2023

Model Overview

This model is a text classification model fine-tuned on the CodeBERT architecture, specifically designed for detecting malicious URLs. It demonstrates moderate accuracy and F1 scores on the evaluation dataset.

Model Features

Fine-tuned on CodeBERT

Utilizes the powerful CodeBERT pre-trained model for fine-tuning, suitable for code-related text analysis

Multi-classification Capability

Capable of classifying URLs into multiple categories to identify different types of malicious URLs

Moderate Performance

Achieved an accuracy of 0.7279 and a weighted F1 score of 0.6508 on the evaluation dataset

Model Capabilities

Malicious URL Detection

Text Classification

Multi-class Classification

Use Cases

Cybersecurity

Malicious URL Filtering

Used to identify and filter potentially malicious URLs

72.79% accuracy

Cybersecurity Analysis

Assists cybersecurity systems in URL risk assessment

🚀 codebert-base-Malicious_URLs

This model is a fine - tuned version of microsoft/codebert-base, designed for text - classification tasks. It offers reliable performance in evaluating malicious URLs, providing multiple metrics to measure its effectiveness.

🚀 Quick Start

This model is a fine - tuned version of microsoft/codebert-base. It achieves the following results on the evaluation set:

Loss: 0.8225
Accuracy: 0.7279
Weighted f1: 0.6508
Micro f1: 0.7279
Macro f1: 0.4611
Weighted recall: 0.7279
Micro recall: 0.7279
Macro recall: 0.4422
Weighted precision: 0.6256
Micro precision: 0.7279
Macro precision: 0.5436

📚 Documentation

✨ Features

Fine - tuned Model: Based on the microsoft/codebert-base, it is optimized for malicious URL classification.
Multiple Metrics: Provides a comprehensive evaluation through various metrics such as accuracy, f1, recall, and precision.

🔧 Technical Details

Model description

For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Multiclass%20Classification/Malicious%20URLs/Malicious%20URLs%20-%20CodeBERT.ipynb

Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

Training and evaluation data

Dataset Source: https://www.kaggle.com/datasets/sid321axn/malicious-urls-dataset

Input Word Length:

Length of Input Text (in Words)

Input Word Length By Class:

Length of Input Text (in Words) By Class

Class Distribution:

Length of Input Text (in Words)

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e - 05
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Weighted f1	Micro f1	Macro f1	Weighted recall	Micro recall	Macro recall	Weighted precision	Micro precision	Macro precision
0.8273	1.0	6450	0.8225	0.7279	0.6508	0.7279	0.4611	0.7279	0.7279	0.4422	0.6256	0.7279	0.5436

Framework versions

Transformers 4.27.4
Pytorch 2.0.0
Datasets 2.11.0
Tokenizers 0.13.3

Property	Details
Model Type	Fine - tuned version of microsoft/codebert - base
Training Data	https://www.kaggle.com/datasets/sid321axn/malicious-urls-dataset

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご