forked from hughec/yelp_social_sentiment
-
Notifications
You must be signed in to change notification settings - Fork 0
/
classifyyelp.py
50 lines (41 loc) · 1.68 KB
/
classifyyelp.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
#!/usr/bin/env python
import readyelp
import cleanyelp
import baselineclassifier
import reviewcrf
import reviewgraph
from sklearn import metrics
from sklearn.feature_extraction.text import TfidfVectorizer
def main():
## Only call the below once, when data needs to be cleaned and split ##
cleanyelp.split_data_by_business(0.75)
#######################################################################
train_reviews = readyelp.read_reviews_to_dict("./train_reviews.json")
test_reviews = readyelp.read_reviews_to_dict("./test_reviews.json")
user_dict = readyelp.read_users_to_dict("./users_limited.json")
klass_list = ["negative", "positive"]
# Calculate class preferences of individual classifier
ind_pref = baselineclassifier.bag_of_words_probabilities(train_reviews, test_reviews)
print "Individual preferences calculated."
# Train CRF model
reviewcrf.train_crf(train_reviews, user_dict)
# Calculate pair strengths
pair_str = reviewcrf.crftag_probabilities(test_reviews, user_dict)
print "Pair strengths calculated."
# Build review graph
min_cut_classes = reviewgraph.build_graph(klass_list, test_reviews, ind_pref, pair_str)
print "Graph cut."
# Make min-cut classification
Y_gold = []
Y_predict = []
for test_id in test_reviews:
review = test_reviews[test_id]
Y_gold.append(review["rating"])
if min_cut_classes[test_id] == 1:
Y_predict.append("positive")
else:
Y_predict.append("negative")
classification_metrics = metrics.classification_report(Y_gold, Y_predict, target_names = klass_list)
print classification_metrics
if __name__ == "__main__":
main()