# BERT FINETUNED ON PHISHING DETECTION

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),
capable of detecting 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

## 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.



## 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 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              |