/
senti_analysis.py
209 lines (160 loc) · 7.33 KB
/
senti_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
from nltk.featstruct import FeatStruct, FeatList, FeatDict
import nltk, os, logging, json, ConfigParser, codecs
import nltk.classify.util
from nltk.classify import NaiveBayesClassifier
class sample_nltk:
def __init__(self):
self.config = ConfigParser.ConfigParser()
self.config.read("senti_analysis.config")
#File names
self.bag_of_words_file_dir = self.config.get('GLOBAL','bag_of_words_file_dir')
self.bag_of_words_file_name = self.config.get('GLOBAL', 'bag_of_words_file_name')
self.feat_file_dir = self.config.get('GLOBAL', 'feat_file_dir')
self.feat_file_name = self.config.get('GLOBAL', 'feat_file_name')
self.test_feat_file_dir = self.config.get('GLOBAL', 'test_feat_file_dir')
self.test_feat_file_name = self.config.get('GLOBAL', 'test_feat_file_name')
self.preprocessing_results_dir = self.config.get('PRE_PROCESSING', 'pre_processing_results_dir')
self.preprocessing_results_file_name = self.config.get('PRE_PROCESSING', 'pre_processing_results_file_name')
self.preprocessing_results_file = os.path.join(self.preprocessing_results_dir, self.preprocessing_results_file_name)
self.logger_file = os.path.join("OUTPUT", "senti_analysis.log")
#Global DS
self.id_to_word_dict = {}
self.id_to_bow_dict = {}
self.tokenized_without_stop_words_list = []
self.stemmed_word_list = []
self.stop_words_set = set()
def run_main(self):
self.preprocessing()
self.feature_selection()
self.feature_extraction()
self.classification()
def preprocessing(self):
self.initialize_logger()
self.parse_reviews()
#self.remove_stop_words()
#self.do_stemming()
#self.dump_preprocessing()
def feature_selection(self):
#Collect positive and negative reviews
self.pos_reviews, self.neg_reviews = self.get_labelled_reviews(self.train_reviews_list)
self.test_pos_reviews, self.test_neg_reviews = self.get_labelled_reviews(self.test_reviews_list)
#Format it acc to classifier input
self.pos_tagged_reviews, self.neg_tagged_reviews = self.tag_words_with_labels(self.pos_reviews, self.neg_reviews)
self.test_pos_tagged_reviews, self.test_neg_tagged_reviews = self.tag_words_with_labels(self.test_pos_reviews, self.test_neg_reviews)
#Since polarity is labelled, we can combine both positive and negative reviews
self.train_reviews = self.pos_tagged_reviews + self.neg_tagged_reviews
self.test_reviews = self.test_pos_tagged_reviews + self.test_neg_tagged_reviews
def feature_extraction(self):
pass
def classification(self):
fstruct = FeatStruct(self.train_reviews)
classifier = NaiveBayesClassifier.train(fstruct)
print 'accuracy:', nltk.classify.util.accuracy(classifier, self.test_reviews)
classifier.show_most_informative_features()
def get_labelled_reviews(self, reviews_list):
pos_reviews_list = neg_reviews_list = []
for review in reviews_list:
words_list = []
if not review:
continue
label = review[0]
if int(label) >= 7:
label = True
elif int(label) <=4:
label = False
else:
label = False
if label:
for word_freq_pair in review.split(" ")[1:]:
if not word_freq_pair:
continue
word_id = word_freq_pair.split(":")[0]
word = self.id_to_word_dict.get(int(word_id))
if not word:
continue
words_list.append(word)
pos_reviews_list.append(words_list)
else:
for word_freq_pair in review.split(" ")[1:]:
if not word_freq_pair:
continue
word_id = word_freq_pair.split(":")[0]
word = self.id_to_word_dict.get(int(word_id))
if not word:
continue
words_list.append(word)
neg_reviews_list.append(words_list)
return (pos_reviews_list, neg_reviews_list)
def tag_words_with_labels(self, pos_reviews, neg_reviews):
pos_tagged_words = []
neg_tagged_words = []
for review in pos_reviews:
word_dict = self.tag_words(review)
if not word_dict:
continue
pos_tagged_words.append((word_dict, 'pos'))
for review in neg_reviews:
neg_word_dict = self.tag_words(review)
if not neg_word_dict:
continue
neg_tagged_words.append((neg_word_dict, 'neg'))
return(pos_tagged_words, neg_tagged_words)
def tag_words(self, words_list):
tag_word_list = []
for word in words_list:
word_tuple = (word, True)
tag_word_list.append(word_tuple)
return dict(tag_word_list)
def initialize_logger(self):
logging.basicConfig(filename=self.logger_file, level=logging.INFO)
logging.info("Initialized logger")
def parse_reviews(self):
self.open_files()
self.load_data()
self.close_files()
def open_files(self):
self.bow = open(os.path.join(self.bag_of_words_file_dir, self.bag_of_words_file_name), 'r')
self.feat = open(os.path.join(self.feat_file_dir, self.feat_file_name), 'r')
self.test_feat = open(os.path.join(self.test_feat_file_dir, self.test_feat_file_name), 'r')
def load_data(self):
self.load_bow()
self.train_reviews_list = self.load_feat(self.feat)
self.test_reviews_list = self.load_feat(self.test_feat)
#self.load_stop_words()
def load_bow(self):
uniq_id = 0
for line in self.bow.readlines():
if not line:
continue
self.id_to_word_dict[uniq_id] = line.strip()
uniq_id += 1
logging.info("id_to_word_dict - length - %s" % (len(self.id_to_word_dict)))
def load_feat(self, fd):
reviews_list = []
for line in fd.readlines():
if not line:
continue
reviews_list.append(line.strip()) #First value in this line hold the labeled value
logging.info("total reviews - length - %s" % (len(reviews_list)))
return reviews_list
def load_stop_words(self):
self.stop_words_set = set(nltk.corpus.stopwords.words())
def close_files(self):
self.bow.close()
self.feat.close()
self.test_feat.close()
def remove_stop_words(self):
for uniq_id, word in self.id_to_word_dict.iteritems():
if word in self.stop_words_set:
self.id_to_word_dict[uniq_id] = '' #Removed stop words will be replaced with null
def do_stemming(self):
stemmer = nltk.stem.PorterStemmer()
for uniq_id, word in self.id_to_word_dict.iteritems():
self.id_to_word_dict[uniq_id] = stemmer.stem(word)
def dump_preprocessing(self):
fd = open(self.preprocessing_results_file, 'w', 'utf-8')
fd.write(json.dumps(self.id_to_word_dict))
fd.close()
if __name__ == "__main__":
sn = sample_nltk()
sn.run_main()