-
Notifications
You must be signed in to change notification settings - Fork 0
/
tweetmanipulation.py
229 lines (181 loc) · 10.4 KB
/
tweetmanipulation.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
import pandas as pd
import sentimenlexicon, ontology
from PIL import Image
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator
import matplotlib.pyplot as plt
import nltk
from collections import Counter
import random
from sklearn import metrics
import utils, sentimenlexicon, sentimenklasifikasisvm
def selectLabelledTweets():
pd.options.mode.chained_assignment = None
# read excel
df = pd.read_excel("data\\Sampel600.xlsx")
# select only the used table
dfClean = df[['class', 'attribute', 'sentimen', 'tweet_asli', 'tweet']]
# remove unlabelled tweet
dfClean['clean_tweet'] = dfClean['tweet_asli'].apply(lambda row: utils.cleanAllTweet(row))
dfClean = dfClean.dropna()
dfClean = dfClean.drop_duplicates()
# dfClean['sentimen_new'] = dfClean.apply(lambda row: sentimen.klasfikasibaru(row), axis=1)
dfClean['attribute_new'] = dfClean['tweet_asli'].apply(lambda row: ontology.attributeClassifier(row))
# dfClean['sentimen_new'] = df.apply(lambda row: sentimen.klasfikasibaru(row), axis=1)
dfClean = dfClean.dropna()
dfClean = dfClean.drop_duplicates()
# save into csv
dfClean.to_csv("data\\tweetclean600-lambda.csv", index_label="id", sep="|")
print(dfClean.shape)
dfOnlyTweeAndClass = dfClean[['sentimen', 'clean_tweet']]
# dfOnlyTweeAndClass = dfClean[['sentimen_new', 'clean_tweet']]
dfOnlyTweeAndClass.to_csv("data\\tweetclean600-only.csv", index_label="id", sep="|")
def infoDataset():
# df = pd.read_csv("data\\tweetclean600-lambda.csv", index_col="id", sep="|")
# df = pd.read_csv("data\\backup\\tweets333.csv", index_col="id", sep="|")
#show best model based 351
# df = pd.read_csv("data\\backup\\tweetclean600-lambda-e916-lk548-ksj351.csv", index_col="id", sep="|")
# sentiment_counts = df.attribute_new.value_counts()
# print("Jumlah distinct : \n"+str(sentiment_counts))
#get describe
df = pd.read_csv("data\\backup\\tweets333-only-withlbscore.csv", index_col="id", sep="|")
print("Describe :")
print(str(df['sentimen_score_lb'].describe()))
# df = pd.read_csv("data\\backup\\tweets333-only-withlbscore-new.csv", index_col="id", sep="|")
# print("Describe :")
# print(str(df.sentimen_new.value_counts()))
def mappingDataset333():
df = pd.read_csv("data\\backup\\tweets333-only-withlbscore.csv", index_col="id", sep="|")
def assignNewClassifier():
df = pd.read_csv("data\\backup\\tweets333-only-withlbscore.csv", index_col="id", sep="|")
mean = df['sentimen_score_lb'].mean()
std = df['sentimen_score_lb'].std()
max = df['sentimen_score_lb'].max()
quartilBawah = -0.004680
quartilAtas = 0.019266
stdper3 = std/3
# rangeBottom = 0 - stdper3
# rangeTop = stdper3
rangeBottom = quartilBawah
rangeTop = quartilAtas
key = "quartil"
print("Nilai "+key+" Range Bottom = "+str(rangeBottom)+" -- Range Top"+str(rangeTop))
df['sentimen_lexicon'] = df['sentimen_score_lb'].apply(lambda row: sentimenlexicon.sentimenLexiconClassifier(row, rangeBottom, rangeTop))
df['sentimen_class_svm_raw'] = sentimenklasifikasisvm.get_classPredictionSVM(list(df['clean_tweet']))
df['sentimen_class_svm'] = df['sentimen_class_svm_raw'].apply(lambda row: sentimenklasifikasisvm.sentimenSVMClassifier(row))
df.to_csv("data\\backup\\tweets333-lexicon-svm-"+key+"-rB"+str(rangeBottom)+"-rT"+str(rangeTop)+".csv", index_label="id", sep="|")
def assignNewClassifierSVM():
df = pd.read_csv("data\\backup\\tweets333-only.csv", index_col="id", sep="|")
df['sentimen_new'] = df['sentimen_score_lb'].apply(lambda row: sentimenlexicon.sentimenLexiconClassifier(row))
df.to_csv("data\\backup\\tweets333-only-withlbscore-new.csv", index_label="id", sep="|")
def separateDataTrainTest():
df = pd.read_csv("data\\tweetclean.csv", sep="|", index_col="id")
# select sentimen netral
dfNetral = df[df['sentimen'] == 'netral']
print("Jumlah sentimen netral = " + str(dfNetral.shape[0]))
# select sentimen positif
dfPositif = df[df['sentimen'] == 'positif']
print("Jumlah sentimen positif = " + str(dfPositif.shape[0]))
# select sentimen negatif
dfNegatif = df[df['sentimen'] == 'negatif']
print("Jumlah sentimen negatif = " + str(dfNegatif.shape[0]))
# pilih 0.7% dan 0.3% dari dataframe netral
dfTrainNetral = dfNetral.iloc[0:int(0.7 * dfNetral.shape[0])]
dfTestNetral = dfNetral.iloc[int(0.7 * dfNetral.shape[0]):dfNetral.shape[0]]
print(str(dfTrainNetral.shape[0]) + "," + str(dfTestNetral.shape[0]))
# pilih 0.7% dan 0.3% dari dataframe positif
dfTrainPositif = dfPositif.iloc[0:int(0.7 * dfPositif.shape[0])]
dfTestPositif = dfPositif.iloc[int(0.7 * dfPositif.shape[0]):dfPositif.shape[0]]
print(str(dfTrainPositif.shape[0]) + "," + str(dfTestPositif.shape[0]))
# pilih 0.7% dan 0.3% dari dataframe netral
dfTrainNegatif = dfNegatif.iloc[0:int(0.7 * dfNegatif.shape[0])]
dfTestNegatif = dfNegatif.iloc[int(0.7 * dfNegatif.shape[0]):dfNegatif.shape[0]]
print(str(dfTrainNegatif.shape[0]) + "," + str(dfTestNegatif.shape[0]))
# Gabung dataframe Train dan Test
dfTrain = pd.concat([dfTrainNetral, dfTrainPositif, dfTrainNegatif])
print("Data train = " + str(dfTrain.shape))
dfTest = pd.concat([dfTestNetral, dfTestPositif, dfTestNegatif])
print("Data tes = " + str(dfTest.shape))
# save datarfame to csv
dfTrain.to_csv("data\\tweetTrain.csv", sep="|", index_label=None)
dfTest.to_csv("data\\tweetTest.csv", sep="|", index_label=None)
def generateWordCloud():
print("Generating stopwords..")
df = pd.read_csv("data\\tweetclean600-only.csv", index_col="id", sep="|")
text = " ".join(tweet for tweet in df.clean_tweet)
stopwords = ["uno","raja salman","coba","beda","program","digaji","sandiaga","mobile","legend","amin","negeri","game","esport","mobile legend","langsung","nggak", "bikin","pernyataan","paham","banget","hati","tusuk","pemerintah", "emang","kali","dukung","bangsa","gajinya","dunia","mati", "semoga","wowo","dipake","april","capres","salah arah", "masyarakat", "pake", "jokowi","presiden","milih","butuh","salah", "ambil","fokus","pilih","negara","beliau", "menang","periode", "negara","ambil","prabowo", "salah arah","udah", "sandi", "indonesia","orang","gak","prabowosandi","jokowimaruf","debat","rakyat","bilang","anak","janji","menghargai","tau","salah arah","mengambil","allah","pemimpin","terpilih"]
wordcloud = WordCloud(stopwords=stopwords,background_color="white").generate(text)
plt.imshow(wordcloud, interpolation='bilinear')
plt.axis("off")
plt.show()
def select350TweetsPerAttribute():
df = pd.read_csv("data\\backup\\tweetclean600-lambda-e916-lk548-ksj351.csv", index_col="id", sep="|")
dfEkonomi = df[df['attribute_new'] == "ekonomi"]
dfEkonomi = dfEkonomi.iloc[0:350]
dfLapanganKerja = df[df['attribute_new'] == "lapangan kerja"]
dfLapanganKerja = dfLapanganKerja.iloc[0:350]
dfKesejahteraan = df[df['attribute_new'] == "kesejahteraan"]
dfKesejahteraan = dfKesejahteraan.iloc[0:350]
allFrames = [dfEkonomi, dfLapanganKerja, dfKesejahteraan]
result = pd.concat(allFrames)
result.to_csv("data\\backup\\tweets333.csv", sep="|", index_label="id")
dfOnlyTweeAndClass = result[['sentimen', 'clean_tweet']]
dfOnlyTweeAndClass.to_csv("data\\backup\\tweets333-only.csv", index_label="id", sep="|")
dfOnlyTweeAndClassAndAttribute = result[['sentimen', 'attribute_new', 'clean_tweet']]
dfOnlyTweeAndClassAndAttribute.to_csv("data\\backup\\tweets333-only-withattribute.csv", index_label="id", sep="|")
def calculateLexiconBasedScore():
df = pd.read_csv("data\\backup\\tweets333-only-withattribute.csv", index_col="id", sep="|")
df['sentimen_score_lb'] = df['clean_tweet'].apply(lambda row: sentimenlexicon.get_sentimen(row))
df.to_csv("data\\backup\\tweets333-only-withattribute-withlbscore.csv", sep="|", index_label="id")
if __name__ == '__main__':
#pilih tweet yg sudah dilabeli
# selectLabelledTweets()
# generate wordcloud
generateWordCloud()
#pisahkan dataset untuk training dan testing
# separateDataTrainTest()
#get info dataset
# infoDataset()
#assignClassifier
# assignNewClassifier()
#select350tweets
# select350TweetsPerAttribute()
#calculate sentimen score lexicon based
# calculateLexiconBasedScore()
#tes
# tweet = "pengamat ekonomi prestasi angka kemiskinan era jokowi terendah sejarah"
# tweet = sentimen.cleanAllTweet(tweet)
# listToken = nltk.word_tokenize(tweet)
# counterToken = Counter(listToken)
#
# counterKesejahteraan = counterToken["kesejahteraan"] + counterToken["miskin"] + counterToken["kemiskinan"] + \
# counterToken["kaya"] + counterToken["kekayaan"] + counterToken["harta"] + counterToken[
# "aset"] + counterToken["biaya hidup"] + counterToken["biaya"] + counterToken["daya beli"]
# counterEkonomi = counterToken["ekonomi"] + counterToken["harga"] + counterToken["inflasi"] + counterToken["pajak"] + \
# counterToken["pertumbuhan"] + counterToken["terjangkau"] + counterToken["murah"] + counterToken[
# "mahal"]
# counterLapanganKerja = counterToken["lapangan kerja"] + counterToken["pekerjaan"] + counterToken["pengangguran"] + \
# counterToken["nganggur"] + counterToken["phk"] + counterToken["gaji"] + counterToken[
# "penghasilan"] + counterToken["tunjangan"]
# counterAll = [counterEkonomi, counterKesejahteraan, counterLapanganKerja]
# print("List = "+str(counterAll))
# maxCount = max(counterAll)
# print("Max = "+str(maxCount))
# indexPosition = counterAll.index(maxCount)
# print("Index max = "+str(indexPosition))
#
# attEkonomi = "ekonomi"
# attKesejahteraan = "kesejahteraan"
# attLapanganKerja = "lapangan kerja"
# listAttribute = [attEkonomi, attKesejahteraan, attLapanganKerja]
# listAttributeEkoLK = [attEkonomi, attLapanganKerja]
# listAttributeKsjLK = [attKesejahteraan, attLapanganKerja]
#
# if (counterEkonomi == counterKesejahteraan):
# print("Masuk")
# print(str(listAttribute[random.randint(0, 1)]))
# elif (counterEkonomi == counterLapanganKerja):
# print(str(listAttributeEkoLK[random.randint(0, 1)]))
# elif (counterKesejahteraan == counterLapanganKerja):
# print(str(listAttributeKsjLK[random.randint(0, 1)]))
# else:
# print(str(listAttribute[indexPosition]))