TwitterSentimentAnalysis/classifier.py at master · stamaimer/TwitterSentimentAnalysis · GitHub

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
import collections
import nltk
import util

icons = [u'\U0001f600', u'\U0001f601', u'\U0001f602', u'\U0001f603',
         u'\U0001f604', u'\U0001f606', u'\u0001f607', u'\U0001f608',
         u'\U0001f60A', u'\U0001F60B', u'\U0001f60c', u'\U0001f60E',
         u'\U0001f60f', u'\U0001f61b', u'\U0001f61c', u'\U0001f61D',
         u'\U0001f62c', u'\U0001f638', u'\U0001f639', u'\U0001f63A',
         u'\U0001f63C', u'\U0001f642', u'\U0001f648', u'\U0001f649',
         u'\U0001f64B', u'\U0001f60D', u'\U0001f617', u'\U0001f618',
         u'\U0001f619', u'\U0001f61a', u'\U0001f63b', u'\U0001f63d',
         u'\U0001f646', u'\U0001f647', u'\U0001f64f', u'\U0001f613',
         u'\U0001f61f', u'\U0001f628', u'\U0001f630', u'\U0001f631',
         u'\U0001f64a', u'\U0001f605', u'\U0001f62e', u'\U0001f62f',
         u'\U0001f632', u'\U0001f633', u'\U00016f35', u'\U0001f640',
         u'\U0001f614', u'\U0001f616', u'\U0001f61e', u'\U0001f622',
         u'\U0001f623', u'\U0001f625', u'\U0001f626', u'\U0001f627',
         u'\U0001f629', u'\U0001f62b', u'\U0001f62d', u'\U0001f63f',
         u'\U0001f641', u'\U0001f64d', u'\U0001f64e', u'\U0001f610',
         u'\U0001f611', u'\U0001f612', u'\U0001f615', u'\U0001f645',
         u'\U0001f620', u'\U0001f621', u'\U0001f624', u'\U0001f63e',
         u'\U0001f609']


deal = [1, util.Tokenizer().tokenize, nltk.bigrams, nltk.trigrams]

def get_ngrams(twitter, n):

    if 1 == n:

        collection = twitter.unigram

        records = collection.find({'frequency':{'$gte':1460, '$lte':107699}}, {'collocation':1, '_id':0})

        print 'generate unigram'

    elif 2 == n:

        collection = twitter.bigram

        records = collection.find({'frequency':{'$gte':75, '$lte':98}}, {'collocation':1, '_id':0})

        print 'generate bigram'

    elif 3 == n:

        collection = twitter.trigram

        records = collection.find({'frequency':{'$gte':2, '$lte':2}}, {'collocation':1, '_id':0})

        print 'generate trigram'

    collocations = []

    for record in records:

        if record['collocation'] not in icons:

            collocations.append(record['collocation'])

    return collocations

def gen_featvect(twitter, n):

    collocations = get_ngrams(twitter, n)

    labeled = twitter.labeled

    count = labeled.count()

    labeled = labeled.find({'label':{'$ne':0}}, {'text':1, 'label':1, '_id':0})

    featvect = []

    print 'generate featvect'

    for item in labeled:

        featdist = {}

        if 1 == n:

            results = deal[n](item['text'])

        else:

            results = deal[n](deal[1](item['text']))

        for collocation in collocations:

            if collocation in results:

                featdist[collocation] = 1

            else:

                featdist[collocation] = 0

        tmp = (featdist, item['label'])

        featvect.append(tmp)

    return featvect

def naive_bayes_classifier(twitter, n):

    featvect = gen_featvect(twitter, n)

    for i in range(3):

        testset = featvect[124167 * i : 124167 * (i + 1)]

        del featvect[124167 * i : 124167 * (i + 1)]

        trainset = featvect

        classifier = nltk.NaiveBayesClassifier.train(trainset)

        print 'accuracy of naivebayes classifier : ', nltk.classify.accuracy(classifier, testset)

	rset = collections.defaultdict(set)
	tset = collections.defaultdict(set)

	for n, (feature, classification) in enumerate(testsets):

            rset[classification].add(n)

            temp = classifier.classify(feature)

            tset[temp].add(n)

	print "precision of 1 : ", nltk.metrics.precision(rset[1], tset[1]), "\trecall of 1 : ", nltk.metrics.recall(rset[1], tset[1]), "\tf_measure of 1 : ", nltk.metrics.f_measure(rset[1], tset[1])
	print "precision of 2 : ", nltk.metrics.precision(rset[2], tset[2]), "\trecall of 2 : ", nltk.metrics.recall(rset[2], tset[2]), "\tf_measure of 2 : ", nltk.metrics.f_measure(rset[2], tset[2])
	print "precision of 3 : ", nltk.metrics.precision(rset[3], tset[3]), "\trecall of 3 : ", nltk.metrics.recall(rset[3], tset[3]), "\tf_measure of 3 : ", nltk.metrics.f_measure(rset[3], tset[3])
	print "precision of 4 : ", nltk.metrics.precision(rset[4], tset[4]), "\trecall of 4 : ", nltk.metrics.recall(rset[4], tset[4]), "\tf_measure of 4 : ", nltk.metrics.f_measure(rset[4], tset[4])
	print "precision of 5 : ", nltk.metrics.precision(rset[5], tset[5]), "\trecall of 5 : ", nltk.metrics.recall(rset[5], tset[5]), "\tf_measure of 5 : ", nltk.metrics.f_measure(rset[5], tset[5])
	print "precision of 6 : ", nltk.metrics.precision(rset[6], tset[6]), "\trecall of 6 : ", nltk.metrics.recall(rset[6], tset[6]), "\tf_measure of 6 : ", nltk.metrics.f_measure(rset[6], tset[6])
	print "precision of 7 : ", nltk.metrics.precision(rset[7], tset[7]), "\trecall of 7 : ", nltk.metrics.recall(rset[7], tset[7]), "\tf_measure of 7 : ", nltk.metrics.f_measure(rset[7], tset[7])
	print "precision of 8 : ", nltk.metrics.precision(rset[8], tset[8]), "\trecall of 8 : ", nltk.metrics.recall(rset[8], tset[8]), "\tf_measure of 8 : ", nltk.metrics.f_measure(rset[8], tset[8])

	print "\n======================================================================\n"

	#print classifier.show_most_informative_features(5)


def max_ent_classifier(twitter, n):

    featvect = gen_featvect(twitter, n)

    for i in range(3):

        testset = featvect[124167 * i : 124167 * (i * 1)]

        del featvect[124167 * i : 124167 * (i + 1)]

        trainset = featvect

        classifier = nltk.MaxentClassifier.train(trainset)

        print 'accuracy of max_ent_classifier : ', nltk.classify.accuracy(classifier, testset)

def svm_classifier(twitter, n):

    featvect = gen_featvect(twitter, n)

    for i in range(3):

        testset = featvect[124167 * i : 124167 * (i * 1)]

        del featvect[124167 * i : 124167 * (i + 1)]

        trainset = featvect

        classifier = nltk.SvmClassifier.train(trainset)

        print 'accuracy of svm_classifier : ', nltk.classify.accuracy(classifier, testset)

def scikit_classifier(twitter, n):

    featvect = gen_featvect(twitter, n)

    for i in range(3):

        testset = featvect[124167 * i : 124167 * (i * 1)]

        del featvect[124167 * i : 124167 * (i + 1)]

        trainset = featvect

        classifier = nltk.SklearnClassifier.train(trainset)

        print 'accuracy of sklearn_classifier : ', nltk.classify.accuracy(classifier, testset)