Update dga-detection.py
Browse files- dga-detection.py +11 -1
dga-detection.py
CHANGED
@@ -2,12 +2,22 @@ import datasets
|
|
2 |
import pandas as pd
|
3 |
import os
|
4 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
class MyDataset(datasets.GeneratorBasedBuilder):
|
7 |
def _info(self):
|
8 |
|
9 |
return datasets.DatasetInfo(
|
10 |
-
description=
|
11 |
features=datasets.Features(
|
12 |
{"domain": datasets.Value("string"), "label": datasets.Value("string")}
|
13 |
),
|
|
|
2 |
import pandas as pd
|
3 |
import os
|
4 |
|
5 |
+
_DESCRIPTION = """\
|
6 |
+
A dataset containing both DGA and normal domain names. The normal domain names were taken from the Alexa top one million domains. An additional 3,161 normal
|
7 |
+
domains were included in the dataset, provided by the Bambenek Consulting feed. This later group is particularly interesting since it consists of suspicious domain
|
8 |
+
names that were not generated by DGA. Therefore, the total amount of domains normal in the dataset is 1,003,161. DGA domains were obtained from the repositories
|
9 |
+
of DGA domains of Andrey Abakumov and John Bambenek. The total amount of DGA domains is 1,915,335, and they correspond to 51 different malware families. DGA domains
|
10 |
+
were generated by 51 different malware families. About the 55% of of the DGA portion of dataset is composed of samples from the Banjori, Post, Timba, Cryptolocker,
|
11 |
+
Ramdo and Conficker malware.
|
12 |
+
"""
|
13 |
+
|
14 |
+
_HOMEPAGE = "https://https://huggingface.co/datasets/harpomaxx/dga-detection"
|
15 |
|
16 |
class MyDataset(datasets.GeneratorBasedBuilder):
|
17 |
def _info(self):
|
18 |
|
19 |
return datasets.DatasetInfo(
|
20 |
+
description=_DESCRIPTION,
|
21 |
features=datasets.Features(
|
22 |
{"domain": datasets.Value("string"), "label": datasets.Value("string")}
|
23 |
),
|