yuudai-g.github.io

The data used to identify attacks and generate rules can be downloaded below.

Training data for machine learning

This data has 31,097 rows and consists of two columns: the URL and the label.

Test data for machine learning (known attacks)

This data has 5,718 rows and consists of two columns: the URL and the label.

Test data for machine learning (variant attacks)

This data has 1,480 rows and consists of two columns: the URL and the label.

10 attack patterns used for rule generation

This data consists only of attack patterns for use in rule generation programs.

161 attack patterns used to compare conventional and proposed methods

This data consists only of attack patterns for use in experiments comparing conventional and proposed methods. IP addresses and DNS names are masked.