TristanBilot/phishGNN

Raw Dataset

asparamancer opened this issue · 7 comments

asparamancer commented

I can see a few references to the file path, but can't see if you provided the raw csv you used - just the processed files?

Do you provide the raw CSV?

Thanks,

TristanBilot commented

Hi,
An example of raw csv is proposed at the end of the dataset presentation: https://tristanbilot.me/phishgnn
Cheers,

shrugginG commented

Hi TristanBilot, I can't access https://tristanbilot.me/phishgnn ......

TristanBilot commented

Hi, sorry about that. Here is the new link: https://tristanbilot.github.io/#/phishgnn

shrugginG commented

Your reply is so quick! Thank so much! I am also a PhD candidate from China and interest in Phishing website detection, I hope to have more opportunities to communicate and learn with you!

TristanBilot commented

with pleasure :) it's always nice collaborating with other researchers!

  • 👍1
shrugginG commented

I apologize for troubling you again. I noticed that the dataset provided on https://tristanbilot.github.io/#/phishgnn is still in the processed .pt format. I am currently looking to reconstruct phishGNN using DGL, and for this purpose, I need the original CSV (maybe as you mentioned in blog:2333_phishing_11176_benign_crawled.csv )files to complete the construction of DGLDataset. Would it be possible for you to provide the original dataset?

TristanBilot commented

hi there, unfortunately I don't have this file anymore, but you can just load the .pt file and extract its content to a .csv as you want to then load it with DGL.