-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathNLP.py
189 lines (153 loc) · 6.02 KB
/
NLP.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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Apr 13 22:57:59 2018
@author: MyReservoir
"""
#ALDA Project- NLP
import random
import os
import numpy as np
import pandas as pd
import treepredict as dt
os.getcwd()
os.chdir('/Users/MyReservoir/Desktop/CSC 522 PROJECT/')
os.getcwd()
import nltk
from nltk.stem import WordNetLemmatizer, PorterStemmer
from nltk.tokenize import RegexpTokenizer, word_tokenize
from pandas import DataFrame
no_of_records=4000
no_of_trials=10
# ----- FUNCTIONS ------
# Data Preprocessing functions
def nltk_process(sample_text):
#reading an input text file
#input_file = open('/Users/jagadeesh/PycharmProjects/ALDA/testdoc.txt','r')
#sample_text = input_file.read()
#input_file.close()
#word tokenizer, but it includes the punctuation
#tokenized = word_tokenize(sample_text)
#tokenizing words using RegexpTokenizer of nltk, which by default removes all punctuation.
tokenizer = RegexpTokenizer(r'\w+')
tokenized = tokenizer.tokenize(sample_text)
length_input = len(tokenized) #used later in for loop which splits pos_tags
# print(tokenized)
#converted each word to its root word using WordNetLemmatizer of nltk.
lemma = WordNetLemmatizer()
lemma_words = map(lemma.lemmatize, tokenized)
pos_tagged = []
#function for assigning pos_tags
def pos_tagging():
try:
for i in lemma_words:
words = nltk.word_tokenize(i)
for j in words:
tagged = nltk.pos_tag(words)
pos_tagged.append(tagged)
except Exception as e:
print(str(e))
#function call for performing pos_tags
pos_tagging()
#print(pos_tagged)
#splitting the pos_tags into two seperate lists for further processing
list_df = DataFrame.from_records(pos_tagged)
words = []
pos_final = []
for i in range(0, length_input):
for j in range(0,2):
if j == 0:
words.append(list_df[0][i][j])
if j == 1:
pos_final.append(list_df[0][i][j])
#print(words)
#print(pos_final)
#converted as table if required for further processing
pos_table = np.column_stack((words, pos_final))
#pos_table=pd.DataFrame(words,pos_final)
return pos_table
def ads(st):
merge = pd.merge(st, pd.DataFrame(dic), how='left', left_on='Word', right_on='Token', sort=False)
merge = merge.fillna(0)
merge2 = pd.merge(merge, pd.DataFrame(intense), how='left', left_on='Word', right_on='Token', sort=False)
merge2 = merge2.fillna(0)
for i in range(len(merge2.index)-1):
merge2['Polarity_x'][i+1]=merge2['Polarity_x'][i+1]*(1+merge2['Polarity_y'][i])
merge3 = pd.merge(merge2, pd.DataFrame(bucket), how='left', left_on='POS', right_on='POS', sort=False)
merge4 = pd.DataFrame(merge3.groupby(by = ['Bucket']).agg({'Polarity_x' : 'sum'}).transpose())
return merge4
#Classification code
#Define function to split dataset with ratio
def splitDataset(dataset, splitRatio):
trainSize = int(len(dataset) * splitRatio)
trainSet = []
copy = dataset.values.tolist()
while len(trainSet) < trainSize:
index = random.randrange(len(copy))
trainSet.append(copy.pop(index))
return [trainSet, copy]
# Data Preprocessing
data = pd.read_csv('/Users/MyReservoir/Desktop/CSC 522 PROJECT/tweetylabel.csv',encoding ='utf=8')
dic = pd.read_csv('/Users/MyReservoir/Desktop/CSC 522 PROJECT/dic.csv',encoding ='utf=8')
intense = pd.read_csv('/Users/MyReservoir/Desktop/CSC 522 PROJECT/intense.csv',encoding ='utf=8')
bucket = pd.read_csv('/Users/MyReservoir/Desktop/CSC 522 PROJECT/bucket.csv',encoding ='utf=8')
data=data[:no_of_records]
data.columns = ['Tweet','Sentiment']
output_nltk = []
for i in range(len(data)):
sample_text = data['Tweet'][i]
output = nltk_process(sample_text)
output_nltk.append(output)
ads_df = pd.DataFrame({'Adjective':[],'Adverb':[],'Noun':[],'Verb':[]})
for i in range(len(output_nltk)):
nltkout_temp = pd.DataFrame(output_nltk[i])
#nltkout_temp = pd.DataFrame(sample_tweet)
nltkout_temp.columns = ['Word','POS']
nltkout_temp['Word'] = pd.DataFrame(nltkout_temp['Word'].str.lower())
ads_df = ads_df.append(ads(nltkout_temp))
ads_df = ads_df.fillna(0)
ads_df['index'] = range(1, len(ads_df) + 1)
data['index'] = range(1, len(data) + 1)
#data['label'] = data.apply(f,axis=1)
ads_df_final = pd.merge(ads_df, pd.DataFrame(data), how='left', left_on='index', right_on='index', sort=False)
dataset = ads_df_final[['Adjective','Adverb','Noun','Verb','Sentiment']]
############# Splitting the Dataset into Testing and Training Sets ##############
final_acc=0.0
for i in range(no_of_trials):
splitRatio = 0.7
trainingSet, testSet = splitDataset(dataset, splitRatio)
#print(trainingSet)
# print(type(trainingSet))
print('Split {0} rows into train = {1} and test = {2} rows'.format(len(dataset),len(trainingSet),len(testSet)))
############# Model Building ##############
b = dt.buildtree(trainingSet)
dt.drawtree(b,jpeg='treeview.jpg')
#print("original_testset=",testSet)
############# Preparing Testing DataSet ##############
testlabels=[]
for i in range(len(testSet)):
label=testSet[i].pop(-1)
testlabels.append(label)
#print("testSet=",testSet)
#print("testlabels=",testlabels)
############# Classification of Test Records ##############
number = 0
for i in range(len(testSet)):
#print("\ntest_data",testSet[i])
a = dt.classify(testSet[i], b)
#print("a=",a)
max=0
best=""
for key in a.keys():
if a[key]>max:
max=a[key]
best=key
#print("best=",best)
#print("label=",testlabels[i])
if(best == testlabels[i]):
number = number + 1
############# Accuracy Calculations ##############
accuracy = (number/len(testSet))* 100
final_acc+=accuracy
final_acc=final_acc/no_of_trials
print('Accuracy: {0}%'.format(final_acc))