-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdataset.py
executable file
·312 lines (259 loc) · 11.5 KB
/
dataset.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
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
import numpy as np
import random
import torch
import math
from random import shuffle
class Reader:
def __init__(self, args, path):
self.ent2id = dict()
self.rel2id = dict()
self.id2ent = dict()
self.id2rel = dict()
self.h2t = {}
self.t2h = {}
self.num_anomalies = 0
self.triples = []
self.start_batch = 0
self.path = path
self.A = {}
self.read_triples()
# if self.path == args.data_dir_YAGO or self.path == args.data_dir_NELL or self.path == args.data_dir_DBPEDIA:
# self.read_triples_yago3()
# else:
# self.read_triples()
self.triple_ori_set = set(self.triples)
self.num_original_triples = len(self.triples)
self.num_entity = self.num_ent()
self.num_relation = self.num_rel()
print('entity&relation: ', self.num_entity, self.num_relation)
self.bp_triples_label = self.inject_anomaly(args)
self.num_triples_with_anomalies = len(self.bp_triples_label)
self.train_data, self.labels = self.get_data()
self.triples_with_anomalies, self.triples_with_anomalies_labels = self.get_data_test()
# def train_triples(self):
# return self.triples["train"]
#
# def valid_triples(self):
# return self.triples["valid"]
#
# def test_triples(self):
# return self.triples["test"]
# def all_triples(self):
# return self.triples["train"] + self.triples["valid"] + self.triples["test"]
def num_ent(self):
return len(self.ent2id)
def num_rel(self):
return len(self.rel2id)
def get_add_ent_id(self, ent):
if ent in self.ent2id:
ent_id = self.ent2id[ent]
else:
ent_id = len(self.ent2id)
self.ent2id[ent] = ent_id
self.id2ent[ent_id] = ent
return ent_id
def get_add_rel_id(self, rel):
if rel in self.rel2id:
rel_id = self.rel2id[rel]
else:
rel_id = len(self.rel2id)
self.rel2id[rel] = rel_id
self.id2rel[rel_id] = rel
return rel_id
def init_embeddings(self, entity_file, relation_file):
entity_emb, relation_emb = [], []
with open(entity_file) as f:
for line in f:
entity_emb.append([float(val) for val in line.strip().split()])
with open(relation_file) as f:
for line in f:
relation_emb.append([float(val) for val in line.strip().split()])
return np.array(entity_emb, dtype=np.float32), np.array(relation_emb, dtype=np.float32)
def read_triples(self):
print('Read begin!')
for file in ["train", "valid", "test"]:
with open(self.path, "r") as f:
for line in f.readlines():
try:
head, rel, tail = line.strip().split("\t")
except:
print(line)
head_id = self.get_add_ent_id(head)
rel_id = self.get_add_rel_id(rel)
tail_id = self.get_add_ent_id(tail)
self.triples.append((head_id, rel_id, tail_id))
self.A[(head_id, tail_id)] = rel_id
# self.A[head_id][tail_id] = rel_id
# generate h2t
if not head_id in self.h2t.keys():
self.h2t[head_id] = set()
temp = self.h2t[head_id]
temp.add(tail_id)
self.h2t[head_id] = temp
# generate t2h
if not tail_id in self.t2h.keys():
self.t2h[tail_id] = set()
temp = self.t2h[tail_id]
temp.add(head_id)
self.t2h[tail_id] = temp
print("Read end!")
return self.triples
def read_triples_yago3(self):
print('Read begin!')
for file in ["train", "valid", "test"]:
with open(self.path + '/' + file + ".txt", "r", encoding="utf-8") as f:
train = f.readlines()
# train_ = set({})
for i in range(len(train)):
x = train[i].split()
x_ = tuple(x)
head, rel, tail = x_[0], x_[1], x_[2]
head_id = self.get_add_ent_id(head)
rel_id = self.get_add_rel_id(rel)
tail_id = self.get_add_ent_id(tail)
self.triples.append((head_id, rel_id, tail_id))
self.A[(head_id, tail_id)] = rel_id
# self.A[head_id][tail_id] = rel_id
# generate h2t
if not head_id in self.h2t.keys():
self.h2t[head_id] = set()
temp = self.h2t[head_id]
temp.add(tail_id)
self.h2t[head_id] = temp
# generate t2h
if not tail_id in self.t2h.keys():
self.t2h[tail_id] = set()
temp = self.t2h[tail_id]
temp.add(head_id)
self.t2h[tail_id] = temp
del (train)
print("Read end!")
return self.triples
def rand_ent_except(self, ent):
rand_ent = random.randint(1, self.num_ent() - 1)
while rand_ent == ent:
rand_ent = random.randint(1, self.num_ent() - 1)
return rand_ent
def generate_neg_triples(self, pos_triples):
neg_triples = []
for head, rel, tail in pos_triples:
head_or_tail = random.randint(0, 1)
if head_or_tail == 0:
new_head = self.rand_ent_except(head)
neg_triples.append((new_head, rel, tail))
else:
new_tail = self.rand_ent_except(tail)
neg_triples.append((head, rel, new_tail))
return neg_triples
def generate_anomalous_triples(self, pos_triples):
neg_triples = []
for head, rel, tail in pos_triples:
head_or_tail = random.randint(0, 2)
if head_or_tail == 0:
new_head = random.randint(0, self.num_entity - 1)
new_relation = rel
new_tail = tail
# neg_triples.append((new_head, rel, tail))
elif head_or_tail == 1:
new_head = head
new_relation = random.randint(0, self.num_relation - 1)
new_tail = tail
else:
# new_tail = self.rand_ent_except(tail)
# neg_triples.append((head, rel, new_tail))
new_head = head
new_relation = rel
new_tail = random.randint(0, self.num_entity - 1)
anomaly = (new_head, new_relation, new_tail)
while anomaly in self.triple_ori_set:
if head_or_tail == 0:
new_head = random.randint(0, self.num_entity - 1)
new_relation = rel
new_tail = tail
# neg_triples.append((new_head, rel, tail))
elif head_or_tail == 1:
new_head = head
new_relation = random.randint(0, self.num_relation - 1)
new_tail = tail
else:
# new_tail = self.rand_ent_except(tail)
# neg_triples.append((head, rel, new_tail))
new_head = head
new_relation = rel
new_tail = random.randint(0, self.num_entity - 1)
anomaly = (new_head, new_relation, new_tail)
neg_triples.append(anomaly)
return neg_triples
def generate_anomalous_triples_2(self, num_anomaly):
neg_triples = []
for i in range(num_anomaly):
new_head = random.randint(0, self.num_entity - 1)
new_relation = random.randint(0, self.num_relation - 1)
new_tail = random.randint(0, self.num_entity - 1)
anomaly = (new_head, new_relation, new_tail)
while anomaly in self.triple_ori_set:
new_head = random.randint(0, self.num_entity - 1)
new_relation = random.randint(0, self.num_relation - 1)
new_tail = random.randint(0, self.num_entity - 1)
anomaly = (new_head, new_relation, new_tail)
neg_triples.append(anomaly)
return neg_triples
def shred_triples(self, triples):
h_dix = [triples[i][0] for i in range(len(triples))]
r_idx = [triples[i][1] for i in range(len(triples))]
t_idx = [triples[i][2] for i in range(len(triples))]
return h_dix, r_idx, t_idx
def shred_triples_and_labels(self, triples_and_labels):
heads = [triples_and_labels[i][0][0] for i in range(len(triples_and_labels))]
rels = [triples_and_labels[i][0][1] for i in range(len(triples_and_labels))]
tails = [triples_and_labels[i][0][2] for i in range(len(triples_and_labels))]
labels = [triples_and_labels[i][1] for i in range(len(triples_and_labels))]
return heads, rels, tails, labels
def all_triplets(self):
ph_all, pr_all, pt_all = self.shred_triples(self.triples)
nh_all, nr_all, nt_all = self.shred_triples(self.generate_neg_triples(self.triples))
return ph_all, pt_all, nh_all, nt_all, pr_all
def get_data(self):
# bp_triples_label = self.inject_anomaly()
bp_triples_label = self.bp_triples_label
labels = [bp_triples_label[i][1] for i in range(len(bp_triples_label))]
bp_triples = [bp_triples_label[i][0] for i in range(len(bp_triples_label))]
bn_triples = self.generate_anomalous_triples(bp_triples)
all_triples = bp_triples + bn_triples
return self.toarray(all_triples), self.toarray(labels)
def get_data_test(self):
bp_triples_label = self.bp_triples_label
labels = [bp_triples_label[i][1] for i in range(len(bp_triples_label))]
bp_triples = [bp_triples_label[i][0] for i in range(len(bp_triples_label))]
return self.toarray(bp_triples), self.toarray(labels)
def toarray(self, x):
return torch.from_numpy(np.array(list(x)).astype(np.int32))
def inject_anomaly(self, args):
print("Inject anomalies!")
original_triples = self.triples
triple_size = len(original_triples)
self.num_anomalies = int(args.anomaly_ratio * self.num_original_triples)
args.num_anomaly_num = self.num_anomalies
print("###########Inject TOP@K% Anomalies##########")
# if self.isInjectTopK:
# self.num_anomalies = args.num_anomaly_num
# print("###########Inject TOP@K Anomalies##########")
# else:
#
# idx = random.sample(range(0, self.num_original_triples - 1), num_anomalies)
idx = random.sample(range(0, self.num_original_triples - 1), self.num_anomalies // 2)
selected_triples = [original_triples[idx[i]] for i in range(len(idx))]
anomalies = self.generate_anomalous_triples(selected_triples) + self.generate_anomalous_triples_2(self.num_anomalies // 2)
triple_label = [(original_triples[i], 0) for i in range(len(original_triples))]
anomaly_label = [(anomalies[i], 1) for i in range(len(anomalies))]
triple_anomaly_label = triple_label + anomaly_label
shuffle(triple_anomaly_label)
return triple_anomaly_label
#
# dataset = Reader(args.data_dir_FB, "train")
# xxx = dataset.inject_anomaly()
# # print(xxx[0][1])
# # print(xxx[0][0])
# #
# xxx, y, a = dataset.get_data()
# print(xxx[0])