-
Notifications
You must be signed in to change notification settings - Fork 0
/
Trollmeter.py
363 lines (315 loc) · 13.3 KB
/
Trollmeter.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
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
import twitterClientManager
import twitterClientManagerUpdate
import tweet_utils
import pandas as pd
import numpy as np
import re
import json
import warnings
from keras.models import load_model
from keras_preprocessing.sequence import pad_sequences
import importlib
importlib.reload(twitterClientManagerUpdate)
warnings.filterwarnings("ignore")
path = r"C:\Users\gianluca.nogara\Desktop\Repo\precollection_trolls"
lst_keys = [
"id",
"id_str",
"user.screen_name",
"text",
"created_at",
"user.statuses_count",
"user.followers_count",
"user.friends_count",
"user.favourites_count",
"retweeted_status.user.screen_name",
"retweeted_status.user.id",
"retweet_count",
"retweet_status.favorite_count",
"retweet_reply_count",
"retweeted_status.user.statuses_count",
"retweeted_status.user.followers_count",
"retweeted_status.user.friends_count",
"retweeted_status.user.favourites_count",
"mention",
"in_reply_to_screen_name",
"in_reply_to_status_id_str"
]
lst_mapped = [
"id",
"id_str",
"screen_name",
"text",
"created_at",
"user_statuses_count",
"user_followers_count",
"user_friends_count",
"user_favourites_count",
"user_retweeted",
"retweet_id_str",
"retweet_count",
"retweet_favorite_count",
"retweet_reply_count",
"user_retweeted_statuses_count",
"user_retweeted_followers",
"user_retweeted_friends",
"user_retweeted_favourites_count",
"mention",
"in_reply_to_screen_name",
"in_reply_to_status_id_str"
]
loaded_model = load_model(path + r"\classifier_trajectory_200.h5")
def refactor_columns(df):
for i in lst_keys:
if i in list(df.columns):
continue
else:
df[i] = [np.NaN for i in range(len(df))]
df = df[lst_keys]
df.columns = lst_mapped
return df
def parse_output(dataframe):
if not "'result_count': 0" in str(dataframe[-1:][0]):
json_data = [r._json for r in dataframe]
df = pd.json_normalize(json_data)
df.rename(columns={"full_text": "text"}, inplace=True)
else:
df = pd.DataFrame()
return df
def parse_output_v2(dataframe):
if len(dataframe) > 0:
json_data = [r._json for r in dataframe]
df = pd.json_normalize(json_data)
else:
df = pd.DataFrame()
return df
def get_keys():
jsonFile = open(path + r'\auth.json', 'r')
config = json.load(jsonFile)
jsonFile.close()
consumer_keys = []
consumer_secrets = []
access_tokens = []
access_token_secrets = []
bearer_tokens = []
for i in config["DEFAULT"]["twitter_credentials"]:
consumer_keys.append(i["consumer_key"])
consumer_secrets.append(i["consumer_secret"])
access_tokens.append(i["access_token"])
access_token_secrets.append(i["access_token_secret"])
bearer_tokens.append(i["bearer_token"])
apis = [consumer_keys, consumer_secrets, access_tokens, access_token_secrets, bearer_tokens]
return apis
def collect_data_update(username):
# print("Starting collection...")
api = get_keys()
twitter_client_v1 = twitterClientManagerUpdate.TwitterClientV1(apis=api, twitter_user=username)
idd, original, reply, retweet, mentions = twitter_client_v1.get_active_interactions()
# print(f"Original: ", len(original))
# print(f"Reply: ", len(reply))
# print(f"Retweet: ", len(retweet))
# print(f"Mentions: ", len(mentions))
rts = twitter_client_v1.get_passive_rts(idd)
# print(f"RTs: ", len(rts))
rps = twitter_client_v1.get_passive_rps(idd)
# print(f"RPs: ", len(rps))
mnts = twitter_client_v1.get_passive_mentions(idd)
# print(f"MNTs: ", len(mnts))
# print("Data collected!")
# print("-----------------")
return original, mentions, reply, retweet, rts, rps, mnts
def collect_data(username):
print("Starting collection...")
api = get_keys()
tweets_number = 10
retweets_number = 5
mention_number = 10
reply_number = 10
user = username
date = ""
tweets = []
retweets = []
mentions = []
replies = []
tweets_ids = []
replies_ids = []
mentions_ids = []
twitter_client_v1 = twitterClientManager.TwitterClientV1(api)
twitter_client_v2 = twitterClientManager.TwitterClientV2(api, user)
print("Collecting tweets for ", user)
tweets.extend(twitter_client_v2.get_tweets(tweets_number))
if not "'result_count': 0" in str(tweets[-1:][0]):
for i in tweets[0]:
tweets_ids.append(str(i.keys).split("=")[1].split(" ")[0])
tweets = twitter_client_v1.get_tweet_by_id(tweets_ids)
else:
print("No tweets...")
print("Collecting retweets received for ", user)
for i in tweets:
twitter_client_v1 = twitterClientManager.TwitterClientV1(api, None, i.id)
retweets.extend(twitter_client_v1.get_retweets_of_tweets(retweets_number))
date = i.created_at
if len(retweets) == 0:
print("No retweets...")
print("Collecting mentions received for ", user)
twitter_client_v2 = twitterClientManager.TwitterClientV2(api, user)
mentions.extend(twitter_client_v2.get_mentions(date, mention_number))
if not "'result_count': 0" in str(mentions[-1:][0]):
for i in mentions[0]:
mentions_ids.append(str(i.keys).split("=")[1].split(" ")[0])
mentions = twitter_client_v1.get_tweet_by_id(mentions_ids)
else:
print("No mentions...")
print("Collecting replies received for", user)
for i in tweets:
twitter_client_v2 = twitterClientManager.TwitterClientV2(api, None, i.id)
replies.extend(twitter_client_v2.get_replies(reply_number))
if not "'result_count': 0" in str(mentions[-1:][0]):
for i in replies[0]:
replies_ids.append(str(i.keys).split("=")[1].split(" ")[0])
replies = twitter_client_v1.get_tweet_by_id(replies_ids)
else:
print("No replies")
print("Data collected!")
print("-----------------")
return tweets, retweets, mentions, replies
def dataset_split(df_activities, df_mentions, df_rps, df_rts, username):
original = df_activities[(df_activities["retweet_id_str"].isna()) &
(df_activities["in_reply_to_status_id_str"].isna())]
for i, row in original.iterrows():
result = re.findall("@([a-zA-Z0-9]{1,15})", row.text)
if len(result) > 0:
original.at[i, "mention"] = result[0]
for i, row in df_mentions.iterrows():
df_mentions.at[i, "mention"] = username
df_tw_a = df_activities[
(df_activities["retweet_id_str"].isna()) &
(df_activities["in_reply_to_status_id_str"].isna()) &
(df_activities["mention"].isna())]
df_m_a = df_activities[df_activities["mention"].notna()]
df_rp_a = df_activities[df_activities["in_reply_to_status_id_str"].notna()]
df_RT_a = df_activities[
(df_activities["in_reply_to_status_id_str"].isna()) &
(df_activities["retweet_id_str"].notna())]
df_RT_m = df_rts
df_rp_m = df_rps
df_m_m = df_mentions
return df_tw_a, df_m_a, df_rp_a, df_RT_a, df_RT_m, df_rp_m, df_m_m
def generate_trajectory(df_tw_a, df_m_a, df_rp_a, df_RT_a, df_RT_m, df_rp_m, df_m_m):
df_author = df_RT_a.append(df_rp_a, ignore_index=True)
df_author = df_author.append(df_tw_a, ignore_index=True)
df_author = df_author.append(df_m_a, ignore_index=True)
df_author.reset_index(inplace=True)
users_subset = pd.unique(df_author.screen_name)
# threshold parameter
t_a = 5 # threshold to filter out trolls (resp. users) that have not performed at least t_a sharing activities
# (original tweets +retweets + replies + mentions)
t_p = 5 # threshold to filter out trolls (resp. users) that have not received at least t_p feedback from other
# accounts (retweets + replies + mentions)
traj = []
user = []
for user_name in users_subset:
# print(f"creating traj for {user_name}")
tw_a_u, n_tw_a = tweet_utils.create_user_dataframe(df_tw_a, user_name, 'screen_name', 1,
"a") # action: original tweet
RT_a_u, n_rt_a = tweet_utils.create_user_dataframe(df_RT_a, user_name, 'screen_name', 2, "a") # action: retweet
rp_a_u, n_rp_a = tweet_utils.create_user_dataframe(df_rp_a, user_name, 'screen_name', 3, "a") # action: reply
m_a_u, n_m_a = tweet_utils.create_user_dataframe(df_m_a, user_name, 'screen_name', 4, "a") # action: mention
n_a = n_tw_a + n_rt_a + n_rp_a + n_m_a # number of actions
# states
RT_p_u, n_rt_p = tweet_utils.create_user_dataframe(df_RT_m, user_name, 'user_retweeted', 5,
"p") # state: got a retweet
rp_p_u, n_rp_p = tweet_utils.create_user_dataframe(df_rp_m, user_name, 'in_reply_to_screen_name', 6,
"p") # state: got a reply
m_p_u, n_m_p = tweet_utils.create_user_dataframe(df_m_m, user_name, 'mention', 7, "p") # state: got a mention
n_p = n_rt_p + n_rp_p + n_m_p # number of states
if n_a > t_a and n_p > t_p: # filtering out some users
# creating actions dataframe
df_active_u = tw_a_u.append(RT_a_u, ignore_index=True)
df_active_u = df_active_u.append(rp_a_u, ignore_index=True)
df_active_u = df_active_u.append(m_a_u, ignore_index=True)
df_active_u = df_active_u.sort_values(by='time')
df_active_u.reset_index(inplace=True)
del df_active_u['index']
df_active_u = tweet_utils.adjust_tweet_df(df_active_u)
# creating states dataframe
RT_p_u = tweet_utils.adjust_retweet_df(RT_p_u)
rp_p_u = tweet_utils.adjust_reply_df(rp_p_u)
m_p_u = tweet_utils.adjust_mention_df(m_p_u)
df_passive_u = RT_p_u.append(rp_p_u, ignore_index=True)
df_passive_u = df_passive_u.append(m_p_u, ignore_index=True)
df_passive_u = df_passive_u.sort_values(by='time')
df_passive_u.reset_index(inplace=True)
df_passive_u = df_passive_u[df_passive_u["tweet_id"].notna()]
del df_passive_u['index']
# create total dataframe
df_total = tweet_utils.merge_df(df_active_u, df_passive_u)
# IRL
trajectories, state_sequence, n_states, n_actions, feature_matrix, t_dict, \
c_dict, traj_lst = tweet_utils.tweet_traj_next_reduced(df_total) # compute trajectories and other
traj.extend(traj_lst)
user.append(user_name)
else:
print("user %s has %s actions and %s states" % (user_name, n_a, n_p))
# print("Traj calculated!")
# print("-----------------")
return pd.DataFrame(list(zip(user, traj)), columns=["screen_name", "state_sequence"])
def reconstruct_traj(row):
f_i = row
f_i = f_i[1:]
f_i = f_i[:-1]
s = str(f_i)
s = s.replace(" ", "")
x = s.split("[")
t = []
for i in range(0, len(x)):
r = x[i].replace("]", "").split(",")
if len(r) >= 2:
t.append([r[0], r[1]])
return t
def remove_action(traj):
seq = []
for i in range(len(traj)):
seq.append(int(traj[i][0]))
return seq
def predict(sequence):
max_length = 200
sequence = pad_sequences(sequence, maxlen=max_length, padding='post')
sequence = sequence.reshape(sequence.shape[0], sequence.shape[1], 1)
return loaded_model.predict(sequence)
def calculate_troll_score(username, collection=True, df_tw_a=None, df_m_a=None, df_rp_a=None, df_RT_a=None,
df_RT_m=None, df_rp_m=None, df_m_m=None):
if collection:
df_tw_a, df_m_a, df_rp_a, df_RT_a, df_RT_m, df_rp_m, df_m_m = collect_data_update(username)
# df_activities = parse_output(tweets)
# df_rts = parse_output_v2(retweets)
# df_mentions = parse_output(mentions)
# df_rps = parse_output(replies)
# df_activities = refactor_columns(df_activities)
# df_mentions = refactor_columns(df_mentions)
# df_rps = refactor_columns(df_rps)
# df_rts = refactor_columns(df_rts)
# df_tw_a, df_m_a, df_rp_a, df_RT_a, df_RT_m, df_rp_m, df_m_m = dataset_split(df_activities, df_mentions,
# df_rps, df_rts, username)
df_tw_a = refactor_columns(df_tw_a)
df_m_a = refactor_columns(df_m_a)
df_rp_a = refactor_columns(df_rp_a)
df_RT_a = refactor_columns(df_RT_a)
df_RT_m = refactor_columns(df_RT_m)
df_rp_m = refactor_columns(df_rp_m)
df_m_m = refactor_columns(df_m_m)
df_traj = generate_trajectory(df_tw_a, df_m_a, df_rp_a, df_RT_a, df_RT_m, df_rp_m, df_m_m)
if len(df_traj) == 0:
print(f"{username} doesn't have enough interaction, cannot calculate Troll score")
return None
else:
df_traj['state_sequence'] = df_traj.apply(lambda row: reconstruct_traj(row['state_sequence']), axis=1)
df_traj['sequence_numbers'] = df_traj.apply(lambda row: remove_action(row['state_sequence']), axis=1)
# print(df_traj["sequence_numbers"]
names = list(df_traj["screen_name"])
lst = predict(df_traj["sequence_numbers"])
for i in range(len(names)):
print(names[i])
print(lst[i][0])
print("_________")
# return predict(df_traj["sequence_numbers"])[0][0]