khuzd/phishgan

URL phishing detection using Generative Adversarial Network (GAN)

phishgan's Issues

version of TensorFlow and Keras

version of TensorFlow and Keras
I want to know what version the author debugged. There are many versions now. I have made mistakes in the latest version.

discuss about abnormal URL and SFH

email new analyse clean

Use func_timeout instead of queue for features extraction

Ip testing

Have to modify the regex in ip testing

php script files

Testing php scripts even if it is a link to a php file

popup new analyse clean

create a new phish database with all new features

Clean all seeds

Disable right click new analyse Clean

Some references for URL classification using DL

Hello Pierrick Robic-Butez,

I tried to send you an email written in your paper, but your address didn't exist (I received an error message from your mail server :( ), so I try to contact you from here.

I’m Keiichi Shima, who asked a question in the Cyberhunt2019 session. I am also working on similar topic, so I’m interested in your talk.

The paper I talked in the question was below.

eXpose: A Character-Level Convolutional Neural Network with Embeddings For Detecting Malicious URLs, File Paths and Registry Keys
https://arxiv.org/abs/1702.08568

We implemented the algorithm written in the paper by ourselves and tried to classify URLs we had. And we found that it actually worked pretty good.

Inspired by the above literature, we also proposed similar method in the short paper below.

Classification of URL bitstreams using bag of bytes
https://ieeexplore.ieee.org/document/8401597

I think there are some other similar activities out there.

Anyway, thank you for your interesting presentation.

I'm also interested in running your code on GitHub. I may ask another question to you :)