bert-finetuned-phishing Open-source Model - Freely Deploy to Detect Phishing Attacks on Websites, Emails, etc.

Bert Finetuned Phishing

Developed by ealvaradob

BERT model fine-tuned on phishing datasets, capable of detecting phishing attacks in URLs, emails, SMS, and website scripts

Text Classification

Transformers

EnglishOpen Source License:Apache-2.0 #Phishing Attack Detection #Multi-format Phishing Identification #High-precision BERT

Downloads 4,436

Release Time : 12/20/2023

Model Overview

This model is a fine-tuned version of bert-large-uncased on phishing datasets, capable of detecting the four most common forms of phishing attacks: URLs, emails, SMS, and website scripts.

Model Features

Multi-scenario Phishing Detection

Capable of detecting various forms of phishing attacks including URLs, emails, SMS, and website scripts

High Accuracy

Achieves 97.17% accuracy on the evaluation set with only a 2.49% false positive rate

Powerful Feature Extraction Based on BERT

Utilizes the strong text comprehension capabilities of the BERT-large model for precise classification

Model Capabilities

Phishing URL Identification

Phishing Email Detection

Phishing SMS Analysis

Malicious Script Recognition

Use Cases

Cybersecurity

Enterprise Email Security Gateway

Integrated into enterprise email systems to automatically filter phishing emails

Reduces phishing email pass-through rate by over 97%

Mobile Application Security Protection

Detects SMS and links received by users in mobile applications

Effectively prevents users from clicking malicious links

🚀 BERT FINETUNED ON PHISHING DETECTION

This model is a fine - tuned version of BERT, capable of detecting phishing in URLs, emails, SMS messages, and websites.

🚀 Quick Start

This model is a fine - tuned version of [bert - large - uncased](https://huggingface.co/bert - large - uncased) on an [phishing dataset](https://huggingface.co/datasets/ealvaradob/phishing - dataset). It can detect phishing in its four most common forms: URLs, Emails, SMS messages, and even websites.

It achieves the following results on the evaluation set:

Loss: 0.1953
Accuracy: 0.9717
Precision: 0.9658
Recall: 0.9670
False Positive Rate: 0.0249

✨ Features

Broad Detection Scope: Capable of detecting phishing in URLs, emails, SMS messages, and websites.
High Performance: Achieves high accuracy, precision, and recall on the evaluation set.

📚 Documentation

Model description

BERT is a transformers model pretrained on a large corpus of English data in a self - supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from those texts.

This model has the following configuration:

24 - layer
1024 hidden dimension
16 attention heads
336M parameters

Motivation and Purpose

Phishing is one of the most frequent and most expensive cyber - attacks according to several security reports. This model aims to efficiently and accurately prevent phishing attacks against individuals and organizations. To achieve it, BERT was trained on a diverse and robust dataset containing: URLs, SMS Messages, Emails, and Websites, which allows the model to extend its detection capability beyond the usual and to be used in various contexts.

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e - 05
train_batch_size: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9, 0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	False Positive Rate
0.1487	1.0	3866	0.1454	0.9596	0.9709	0.9320	0.0203
0.0805	2.0	7732	0.1389	0.9691	0.9663	0.9601	0.0243
0.0389	3.0	11598	0.1779	0.9683	0.9778	0.9461	0.0156
0.0091	4.0	15464	0.1953	0.9717	0.9658	0.9670	0.0249

Framework versions

Transformers 4.34.1
Pytorch 2.1.1+cu121
Datasets 2.14.6
Tokenizers 0.14.1

📄 License

This model is licensed under the Apache 2.0 license.

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