The dataset has been prepared from: https://www.kaggle.com/datasets/taruntiwarihp/phishing-site-urls?resource=download. The dataset is balanced dataset which has 2000 each data samples for phising and non-phising mails. It has been divided into three parts - train, validation and test. The label 0 represent that it's a non-phising URL while 1 represents phising.
7fbd14c verified metadata
dataset_info:
features:
- name: text
dtype: string
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 180713
num_examples: 2800
- name: validation
num_bytes: 37814
num_examples: 600
- name: test
num_bytes: 41008
num_examples: 600
download_size: 189016
dataset_size: 259535
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
size_categories:
- 1K<n<10K