-
Notifications
You must be signed in to change notification settings - Fork 0
/
Preprocessing.py
104 lines (81 loc) · 3.52 KB
/
Preprocessing.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
import pickle
import random
import nltk
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
class PreprocessingDataSet:
def __init__(self):
random.seed(1725)
self.word_features = {}
self.stop_words = stopwords.words('english')
# for removing stopping word
def remove_stopping_word(self, raw_data):
filtered_list = []
for w in raw_data:
if not w in self.stop_words:
filtered_list = [w]
return filtered_list
# have to implement this function to get more accuracy in result
def getUniqueItems(self, items):
uniquelist = []
for item in items:
if item not in uniquelist:
uniquelist.append(item)
return uniquelist
def word_to_vector(self, word_features):
return {token: idx for idx, token in enumerate(set(word_features))}
def find_features(self, document):
words = word_tokenize(document)
features = {}
for w in words:
features[w] = (w in self.word_features.keys())
return features
def get_training_testing_dataset(self, training_percentage=0.5):
spathneg = "dataset/negativesenwosc.txt"
spathpos = "dataset/positivesenwosc.txt"
short_pos = open(spathpos, "r", encoding='utf-8').read()
short_neg = open(spathneg, "r", encoding='utf-8').read()
# move this up here
all_words = []
documents = []
# word tag and its meaning J is for adjective,
# V is for verb
allowed_word_types = ["J", "V"]
print("\n70 30 verb adjective")
for p in short_pos.split('\n\n\n'):
documents.append((p, "pos"))
words = word_tokenize(p)
pos = nltk.pos_tag(words)
for w in pos:
if w[1][0] in allowed_word_types:
all_words.append(w[0].lower())
for p in short_neg.split('\n\n\n'):
documents.append((p, "neg"))
words = word_tokenize(p)
pos = nltk.pos_tag(words)
for w in pos:
if w[1][0] in allowed_word_types:
all_words.append(w[0].lower())
# saving documents "no need can be removed"
save_documents = open("pickle_data_model_save/documents.pickle", "wb")
pickle.dump(documents, save_documents)
save_documents.close()
# counting freq of words
all_words = nltk.FreqDist(all_words)
# selecting 5000 word for training
self.word_features = list(all_words.keys())[:5000]
# converting word to vector
self.word_features = self.word_to_vector(self.word_features)
# saving word_features for testing purpose
save_word_features = open("pickle_data_model_save/word_features.pickle", "wb")
pickle.dump(self.word_features, save_word_features)
save_word_features.close()
# prepating feature set with tag for training
featuresets = [(self.find_features(rev), category) for (rev, category) in documents]
# random shuffling for data set trainig for better accuracy
random.shuffle(featuresets)
length = len(featuresets)
# preparing training and testing dataset for 70-30 percentage
training_set = featuresets[:int(length * training_percentage)]
testing_set = featuresets[int(length * training_percentage):]
return (training_set, testing_set)