-
Notifications
You must be signed in to change notification settings - Fork 1
/
solution22.py
328 lines (266 loc) · 14.8 KB
/
solution22.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
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
"""
this program finds the distance between user question and the available questions and answers , in order to find the most matching question
Semantic similarity
1. Sentence Segmentation: we split the input string/question string and answers strings into sentences. and further sentences into words.
2. Abbreviations of acronyms: we expand the captilized possible acronyms with abbreviations given by a file(online).
3. Stop words elimination: we eliminate the stop from the word vectors.
4. Remove Punctuation: we remove punctuations.
5. Parts of speech Tagging: the words are tagged to their corresponding parts of english sentence grammar.
6. Word sense Identification: we find the word sense of the words using wordnet package. it gives the synonyms, and definitions of the word senses
7. Word Sense Disambiguation using Lesk algorithm: we capture the most appropriate word sense applicable from the synonyms provided by wordnet.
8. Path similarity between word senses: word senses are stored in a paths representative of their etymology including hypernyms and hyponyms of the word.Path similarity gives a notion of their closeness.
9. lch_similairty between the word senses: Leacock-Chodorow Similarity: a score denoting how similar two word senses are, based on the shortest path that connects the senses
10. wup_similarity between the word senses: Wu-Palmer Similarity: how similar two word senses are, based on the depth of the two senses in the taxonomy
11. we find the max similarity between any words of two strings and add max similarity for all words in from the strings
Vector similarity:
1. we obtain vectors out of the strings using their term-document-frequency and inverse document-frequency score (tdf-idf)
2. we find mathematical similarity between those two vectors using the following scores
a. Cosine similarity: the cosine distance between vector A and B based on the angle between them.
b. Correlation Coefficient: indicates how positiviley or negatively correlated are the vectors.
c. Bray-Curtis Score: Quantify the compositional dissimilarity between two different sites, based on counts at each site.
d. city-Block: Manhattan score: The distance between two points measured along axes at right angles
e. Sorenson-dice coefficient based on bigrams
f. yule distance between vectors: yule dissimilarity score between two boolean arrays
g. hamming distance between vectors: number used to denote the difference between two binary strings
we combine scores using arbitrary weights to produce a total similarity score.
based on which the appropriate question and the appropriate answer are obtained.
"""
import json
from nltk import WordNetLemmatizer
from nltk.corpus import stopwords
import string
from nltk.corpus import wordnet as wn
from nltk.wsd import lesk
import nltk as nltk
import re
import numpy as np
import numpy.linalg as LA
import scipy.spatial.distance as dist
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
from sklearn.metrics import mutual_info_score
#def
def abbrevate(given):
sentence=given.split()
abbr_dict={}
foldr="C:/Users/Yogesh/Desktop/New"
fil='abbr.txt'
with open(flodr+'/'+fil) as abbr_file:
for line in abbr_file:
line_words=line.split('\t');
#print (line_words)
abbr_dict[line_words[0]]=line_words[1]
for w in range(len(sentence)):
caps=1
for ch in sentence[w]:
if(ch.islower()):
caps=0
if(caps==1 and sentence[w] in abbr_dict):
# print(sentence1[w]+"has caps and "+abbr_dict[sentence1[w]]);
# print(string1.count(sentence1[w]));
given=given.replace(sentence[w],abbr_dict[sentence[w]],1)
return given
def similarity(string1,string2):
#split the string into sentences and sentences into words
sentences1=[d for d in re.split('\.\W',string1)]
sentences2=[d for d in re.split('\.\W',string2)]
#sentences1=[d for d in string1.split('|')]
#sentences2=[d for d in string2.split('|')]
sentences1=[d.split('|') for d in sentences1]
sentences1=[d for e in sentences1 for d in e]
sentences2=[d.split('|') for d in sentences2]
sentences2=[d for e in sentences2 for d in e]
#remove stop words
stop=stopwords.words('english')
#clean_sentence=[ j for j in sentence if j not in stop]
#sentence1=[j for j in sentence1 if j not in stop]
#sentence2=[j for j in sentence2 if j not in stop]
#remove Punctuation
regex=re.compile('[%s]' % re.escape(string.punctuation))
#remove_punct_map=dict((ord(char),None) for char in string.punctuation)
#regex.sub('', s)
#clean_sentence_npunct=[c.translate(remove_punct_map) for c in clean_sentence]
sentences1=[regex.sub('', c) for c in sentences1]
sentences2=[regex.sub('', c) for c in sentences2]
sentences1=[c.replace('\n','') for c in sentences1]
sentences2=[c.replace('\n','') for c in sentences2]
sentences1=[c.replace('\t','') for c in sentences1]
sentences2=[c.replace('\t','') for c in sentences2]
#print(sentences1)
#print(sentences2)
totalsimilarity=0;
for sent1 in sentences1:
sentence1=sent1.split()
for k,ktag in nltk.pos_tag(sentence1):
maxsimilarity=0
syn1=lesk(sentence1,k)
#print(k+" and"+ktag)
if syn1 is None:
#print("syn1 is none");
for i,sym1 in enumerate(wn.synsets(k)):
syn1=sym1
if syn1 is not None:
for sent2 in sentences2:
sentence2=sent2.split()
for j,jtag in nltk.pos_tag(sentence2):
# print(syn1.name()+" and "+syn2.name())
sim=0
a=0
#print(j+ " and " +jtag)
# if(ktag is jtag):
if(a==0):
syn2=lesk(sentence2,j)
#print(syn1.name()+" and "+syn2.name())
if syn2 is None:
#print("syn2 is none")
for i,sym2 in enumerate(wn.synsets(j)):
syn2=sym2
if syn2 is not None:
#print(j+"is none")
#else:
# print(syn1.name()+" and "+syn2.name())
ps=syn1.path_similarity(syn2)
ws=syn1.wup_similarity(syn2)
ls=0
if(syn1.name().split('.')[1] == syn2.name().split('.')[1]):
# print(syn1.name()+" and "+syn2.name())
ls=syn1.lch_similarity(syn2)
if ps is not None:
sim=sim+ps
if ws is not None:
sim=sim+ws
if ls is not None:
sim=sim+ls
if maxsimilarity < sim :
maxsimilarity=sim
totalsimilarity=totalsimilarity+maxsimilarity
#print ("\nscore"+str(totalsimilarity));
return totalsimilarity
data=[]
abbr_dict={}
with open(foldr+"/data.json") as data_file:
for line in data_file:
data.append(json.loads(line))
#"""
#abbreviations data
with open(foldr+"/abbr.txt") as abbr_file:
for line in abbr_file:
line_words=line.split('\t');
#print (line_words)
abbr_dict[line_words[0]]=line_words[1]
#print (line_words)
#"""
#remove stop words
stop=stopwords.words('english')
words=[]
counts=[]
worddict={}
userinput="howb to book a hotel using gocash. is it free?"
#Strinput=Strinput.split(
maxscore=0;
maxquestion=0;
Qvectorizer=CountVectorizer(stop_words=stop)
Avectorizer=CountVectorizer(stop_words=stop)
QTvectorizer=TfidfVectorizer()
ATvectorizer=TfidfVectorizer()
all_questions=[d['question'] for d in data]
all_questions=[userinput]+all_questions
all_answers=[d['answer'] for d in data]
all_answers=[userinput]+all_answers
QuestionTVectorArray=QTvectorizer.fit_transform(all_questions)
AnswerTVectorArray=ATvectorizer.fit_transform(all_answers)
#print "question cosine similairity-->",cosine_similarity(QuestionTVectorArray[0:1],QuestionTVectorArray)
#print "answer cosine similarity-->",cosine_similarity(AnswerTVectorArray[0:1],AnswerTVectorArray)
Qcosines=cosine_similarity(QuestionTVectorArray[0:1],QuestionTVectorArray)
Acosines=cosine_similarity(AnswerTVectorArray[0:1],AnswerTVectorArray)
Qbray=[dist.braycurtis(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Abray=[dist.braycurtis(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qcanberra=[dist.canberra(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Acanberra=[dist.canberra(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qhamming=[dist.hamming(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Ahamming=[dist.hamming(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qcorrelation=[dist.correlation(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Acorrelation=[dist.correlation(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qcityblock=[dist.cityblock(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Acityblock=[dist.cityblock(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qdice=[dist.dice(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Adice=[dist.dice(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
Qyule=[dist.yule(QuestionTVectorArray[0].toarray(),u.toarray()) for u in QuestionTVectorArray]
Ayule=[dist.yule(AnswerTVectorArray[0].toarray(),u.toarray()) for u in AnswerTVectorArray]
#C_Q=np.histogram2d(QuestionTVectorArray[1],QuestionTVectorArray[1])[0]
#print "question mutual info-->",mutual_info_score(None,None,contigency=C_Q)#QuestionTVectorArray[0:1],QuestionTVectorArray)
#QuestionVectorArray=Qvectorizer.fit_transform(all_questions).toarray()
#AnswerVectorArray=Avectorizer.fit_transform(all_answers).toarray()
#QUserinputVectorArray=Qvectorizer.transform(userinput).toarray()
#AUserinputVectorArray=Avectorizer.transform(userinput).toarray()
#cx=lambda a,b:round(np.inner(a,b)/(LA.norm(a)*LA.norm(b)),3)
"""
mincosine=1
minques=0
for Qv in range(len(QuestionVectorArray)):
# print "Question vector:",QuestionVectorArray[Qv]
# print "Answer vector:",AnswerVectorArray[Qv]
# print "input vector:",AUserinputVectorArray[0]
for Quservector in QUserinputVectorArray:
Qcosine=cx(QuestionVectorArray[Qv],Quservector)
Acosine=cx(AnswerVectorArray[Qv],AUserinputVectorArray[0])
print Acosine;
print Qcosine;
sumcosine=Qcosine+Acosine
if(mincosine>sumcosine):
mincosine=sumcosine
minques=Qv
print "best cosine",mincosine,"best question",data[minques]['question'],"best answer",data[minques]['answer']
#"""
for i1 in range(len(data)):
sentence=data[i1]["question"].split()
string1=userinput;
string2=data[i1]["question"]
string3=data[i1]["answer"]
#string3="which is the AAA largest democracy in the world"
string1=abbrevate(string1)
string2=abbrevate(string2)
string3=abbrevate(string3)
#score13=similarity(string1,string3)
# score12=similarity(string1,string2)
questionscore=similarity(string1,string2)
answerscore=similarity(string1,string3)
#print(str(score13)+"is the score")
#print("question score:"+str(questionscore)+" answer score: "+str(answerscore))
totalsimilarity=0
#add similarity scores
totalsimilarity=75*(questionscore)+150*(answerscore)
#subtract bray score,canberra score
totalsimilarity=totalsimilarity+15*(Qbray[i1+1]+Abray[i1+1])+15*(Qcanberra[i1+1]+Acanberra[i1+1])+15*(Qhamming[i1+1]+Ahamming[i1+1])+10*(Qyule[i1+1]+Ayule[i1+1])
# Add Dice score,correlation score,Yule score
totalsimilarity=totalsimilarity+75*(Qcorrelation[i1+1]+Acorrelation[i1+1])+55*(Qcosines[0][i1]+Acosines[0][i1])+50*(Qdice[i1+1]+Adice[i1+1])
print("Total: ",totalsimilarity," cosines:",Qcosines[0][i1]," correlation:",Qcorrelation[i1+1]," bray:",Qbray[i1+1]," canberra:",Qcanberra[i1+1]," hamming:",Qhamming[i1+1]," yule:",Qyule[i1+1]," dice:",Qdice[i1+1])
#sentence1=string1.split()
#sentence2=string2.split()
#sentence3=string3.split()
#print(string1+" and "+string2+" and "+string3
if(maxscore<totalsimilarity):#-15*(Qbray[i1+1]-Abray[i1+1])-15*(Qcanberra[i1+1]+Acanberra[i1+1])-15*(Qhamming[i1+1]+Ahamming[i1+1])-10*(Qyule[i1+1]+Ayule[i1+1])):
maxscore=totalsimilarity#-15*(Qbray[i1+1]-Abray[i1+1])-15*(Qcanberra[i1+1]+Acanberra[i1+1])-15*(Qhamming[i1+1]+Ahamming[i1+1])-10*(Qyule[i1+1]+Ayule[i1+1])
maxquestion=i1
print("\n\n\ngiven question:"+userinput+"\nthe best question\n question:"+data[maxquestion]["question"]+" \nBest answer \nanswer:"+data[maxquestion]["answer"]+"\n score:"+str(maxscore));
"""
#Lemmatize words
Lemmatizer=WordNetLemmatizer()
synonyms=[]
clean_sentence_np_lemmatized=[Lemmatizer.lemmatize(c) for c in clean_sentence_npunct]
clean_sentence_np_lm_related=[]
synonyms=[]
for c in clean_sentence_np_lemmatized:
for i,syn in enumerate(wn.synsets(c)):
synonyms=syn.lemma_names()
synonyms=[n.replace('_','') for n in synonyms]
synonyms=["R-"+n for n in synonyms]
synonyms.append(c)
clean_sentence_np_lm_related=clean_sentence_np_lm_related+synonyms
for k in clean_sentence_np_lm_related:
worddict[k]=clean_sentence_np_lm_related.count(k)
print(clean_sentence)
print(worddict)
print(totalsimilarity)
#"""