-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathbehavior_analysis.py
248 lines (187 loc) · 11.2 KB
/
behavior_analysis.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
__author__ = "Md. Ahsan Ayub"
__license__ = "GPL"
__credits__ = ["Ayub, Md. Ahsan", "Martindale, Nathan", "Smith, Steven",
"Siraj, Ambareen"]
__maintainer__ = "Md. Ahsan Ayub"
__email__ = "[email protected]"
__status__ = "Prototype"
# Importing the libraries
import pandas as pd
import glob
import os
import numpy as np
def identifyDominatingFeatureSpace(aggegated_dataset_flagged, aggegated_dataset_not_flagged):
flagged_items_list = []
non_flagged_items_list = []
# IRP Operations
flagged_items_list.append(round(aggegated_dataset_flagged.numIRP.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numIRP.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.numFSO.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numFSO.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.numFIO.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numFIO.median()))
# Files
flagged_items_list.append(round(aggegated_dataset_flagged.numFileObject.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numFileObject.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.numFileName.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numFileName.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.totalFileName.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.totalFileName.median()))
# IRP Flags
flagged_items_list.append(round(aggegated_dataset_flagged.numIRPFlag.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numIRPFlag.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.numMajorOperationTypes.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numMajorOperationTypes.median()))
# Others
flagged_items_list.append(round(aggegated_dataset_flagged.numStatus.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numStatus.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.numInform.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.numInform.median()))
# Buffer and Entropy
flagged_items_list.append(round(aggegated_dataset_flagged.meanBufferLength.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.meanBufferLength.median()))
flagged_items_list.append(round(aggegated_dataset_flagged.sumEntropy.median()))
non_flagged_items_list.append(round(aggegated_dataset_not_flagged.sumEntropy.median()))
return flagged_items_list, non_flagged_items_list
def benign_dataset_feature_space_exploration(dataset, cols):
temp = [0 * i for i in range(len(cols))]
groupedData = dataset.groupby(['session'])
for item in groupedData:
# IRP Opn
if temp[0] <= round(item[1]['numIRP'].median()):
temp[0] = round(item[1]['numIRP'].median())
# FSO Opn
if temp[1] <= round(item[1]['numFSO'].median()):
temp[1] = round(item[1]['numFSO'].median())
# FIO Opn
if temp[2] <= round(item[1]['numFIO'].median()):
temp[2] = round(item[1]['numFIO'].median())
# File Object
if temp[3] <= round(item[1]['numFileObject'].median()):
temp[3] = round(item[1]['numFileObject'].median())
# Unique Files Accessed
if temp[4] <= round(item[1]['numFileName'].median()):
temp[4] = round(item[1]['numFileName'].median())
# Total File Accessed
if temp[5] <= round(item[1]['totalFileName'].median()):
temp[5] = round(item[1]['totalFileName'].median())
# IRP Flag
if temp[6] <= round(item[1]['numIRPFlag'].median()):
temp[6] = round(item[1]['numIRPFlag'].median())
# Major Opn Type
if temp[7] <= round(item[1]['numMajorOperationTypes'].median()):
temp[7] = round(item[1]['numMajorOperationTypes'].median())
# Status
if temp[8] <= round(item[1]['numStatus'].median()):
temp[8] = round(item[1]['numStatus'].median())
# Inform
if temp[9] <= round(item[1]['numInform'].median()):
temp[9] = round(item[1]['numInform'].median())
# Buffer Length
if temp[10] <= round(item[1]['meanBufferLength'].median()):
temp[10] = round(item[1]['meanBufferLength'].median())
# Entropy
if temp[11] <= round(item[1]['sumEntropy'].median()):
temp[11] = round(item[1]['sumEntropy'].median())
# Machine
temp[12] = round(item[1]['machine'].unique()[0])
# Category
temp[13] = 0
return temp
if __name__ == '__main__':
pwd = os.getcwd()
flag = 1 # 0 for benign and 1 for ransomware
if (flag == 0): # Benign datasets' behavior analysis
os.chdir(pwd)
benign_labels = ["IRP_Opn", "FSO_Opn", "FIO_Opn", "File_Object", "Unique_Files_Accessed", "Total_File_Accessed",
"IRP_Flag", "Major_Opn_Type", "Status", "Inform", "Buffer_Length", "Entropy", "Machine", "Category"]
benign_feature_space_list = []
for index in range(1, 12):
file_path = "./Dataset/benign-irp-logs/machine_" + str(index) + "/"
os.chdir(file_path)
all_filenames = [i for i in glob.glob('*_aggregated.csv')]
all_filenames = sorted(all_filenames)
total_files = len(all_filenames)
for i in range(total_files):
filename = all_filenames[i]
aggregated_dataset = pd.read_csv(filename)
aggregated_dataset = aggregated_dataset.drop(['Unnamed: 0'], axis=1)
print(aggregated_dataset.head())
aggregated_dataset['machine'] = index
aggregated_dataset['session'] = i + 1
aggregated_dataset = aggregated_dataset.drop(['major_opration_type_types_seq_items',
'minor_opration_type_types_seq_items',
'status_seq_items', 'file_name_seq_items'], axis=1)
if (i > 0):
aggregated_dataset_append = aggregated_dataset_append.append(aggregated_dataset)
else:
aggregated_dataset_append = aggregated_dataset # Copy
del aggregated_dataset
benign_feature_space_list.append(benign_dataset_feature_space_exploration(aggregated_dataset_append, benign_labels))
del total_files
del aggregated_dataset_append
os.chdir(pwd)
behavior_benign_feature_space = pd.DataFrame(benign_feature_space_list, columns = benign_labels)
del benign_feature_space_list
# Dump aggregated dataframe
behavior_benign_feature_space.to_csv("./Behavior_Analysis_Results/behaviour_benign_feature_space_exploration.csv")
if (flag == 1): # Ransomware datasets' behavior analysis
os.chdir('./Dataset/ransomware-irp-logs/')
# Storing the file names for all the aggregated datasets
all_filenames = [i for i in glob.glob('*_aggregated*')]
all_filenames = sorted(all_filenames)
# Building lists
labels = ["IRP Opn", "FSO Opn", "FIO Opn", "File Object", "Unique Files Accessed", "Total File Accessed",
"IRP Flag", "Major Opn Type", "Status", "Inform", "Buffer Length", "Entropy",
"Process_Names", "Ransomware_Hash", "Category"]
processAggregateList = []
for filename in all_filenames:
#filename = all_filenames[0]
flagged_items_list = []
non_flagged_items_list = []
try:
aggegated_dataset = pd.read_csv(filename, compression='zip', header=0, sep=',', quotechar='"')
except:
try:
aggegated_dataset = pd.read_csv(filename, compression='gzip', header=0, sep=',', quotechar='"')
except:
try:
aggegated_dataset = pd.read_pickle(filename, compression='gzip')
except:
continue
try:
aggegated_dataset = aggegated_dataset.drop(['Unnamed: 0'], axis=1)
except:
print("No need to drop the column")
print(aggegated_dataset.head())
# Segragating flagged and non flagged dataframes
aggegated_dataset_not_flagged = aggegated_dataset.drop(aggegated_dataset[(aggegated_dataset['doc_files_flag'] != 0)].index)
aggegated_dataset_flagged = aggegated_dataset.drop(aggegated_dataset[(aggegated_dataset['doc_files_flag'] != 1)].index)
# Removing the list items row
aggegated_dataset_flagged = aggegated_dataset_flagged.drop(['major_opration_type_types_seq_items',
'minor_opration_type_types_seq_items',
'status_seq_items', 'file_name_seq_items'], axis=1)
aggegated_dataset_not_flagged = aggegated_dataset_not_flagged.drop(['major_opration_type_types_seq_items',
'minor_opration_type_types_seq_items',
'status_seq_items', 'file_name_seq_items'], axis=1)
flagged_items_list, non_flagged_items_list = identifyDominatingFeatureSpace(aggegated_dataset_flagged,
aggegated_dataset_not_flagged)
# Process names
flagged_items_list.append(aggegated_dataset_flagged.process_name.tolist())
non_flagged_items_list.append(np.NaN)
# Ransomware Hash
filename = filename.split('_')[0]
flagged_items_list.append(filename)
non_flagged_items_list.append(filename)
# Category of the data or family : 0 and 1 indicate benign and malicious respectively
flagged_items_list.append(1)
non_flagged_items_list.append(0)
processAggregateList.append(flagged_items_list)
processAggregateList.append(non_flagged_items_list)
del aggegated_dataset
del aggegated_dataset_flagged
del aggegated_dataset_not_flagged
os.chdir(pwd)
behaviour_feature_space_aggregated = pd.DataFrame(processAggregateList, columns = labels)
# Dump aggregated dataframe
behaviour_feature_space_aggregated.to_csv("./Behavior_Analysis_Results/behaviour_feature_space_aggregated.csv")