def patterns_approach(llimit,ulimit,isphrase,pathname):
pos_pattern_list=[]
neg_pattern_list=[]
trainingdata=[]
trainingclass=[]
pattern_train_set=[]
totalcount=ulimit-llimit
lpcount=0
cnt_var=0
print '\nNo of +ve reviews scanned for training : '
for fid in movie_reviews.fileids(categories=['pos'])[llimit:ulimit]:
test_pattern=proximity_tagger.pattern_list(movie_reviews.abspath(fid),isphrase,cnt_var,0,pathname)
pos_pattern_list.extend(test_pattern)
lpcount=lpcount+1
cnt_var+=1
print 'Scanning +ve review ',lpcount,'.'*10,(float(lpcount)*100/float(totalcount)),'%'
lpcount=0
cnt_var=0
print '\nNo of -ve reviews scanned for training : '
for fid in movie_reviews.fileids(categories=['neg'])[llimit:ulimit]:
test_pattern=proximity_tagger.pattern_list(movie_reviews.abspath(fid),isphrase,cnt_var,1,pathname)
neg_pattern_list.extend(test_pattern)
lpcount=lpcount+1
cnt_var+=1
print 'Scanning -ve review ',lpcount,'.'*10,(float(lpcount)*100/float(totalcount)),'%'
lpcount=0
totalcount=len(pos_pattern_list)
print '\nNo of +ve reviews trained : '
trainingdata.extend(pos_pattern_list)
for i in range(totalcount):
trainingclass.append(1)
lpcount+=1
print 'Training +ve patterns ',lpcount,'.'*10,(float(lpcount)*100/float(totalcount)),'%'
lpcount=0
totalcount=len(neg_pattern_list)
print '\nNo of +ve reviews trained : '
trainingdata.extend(neg_pattern_list)
for i in range(totalcount):
trainingclass.append(0)
lpcount+=1
print 'Training -ve patterns ',lpcount,'.'*10,(float(lpcount)*100/float(totalcount)),'%'
pattern_train_set.append(trainingdata)
pattern_train_set.append(trainingclass)
f = open('train_result\proximity_pattern_train_result_'+str(isphrase),'w')
json.dump(pattern_train_set,f)
f.close()
def review_test(isphrase):
bin_testset=[]
f = open('train_result/proximity_bin_train_result_'+str(isphrase),'r')
bin_testset=json.load(f)
f.close()
#clf=SVC()
clf=LinearSVC()
clf.fit(bin_testset[0],bin_testset[1])
pattern_testset=[]
f = open('train_result/proximity_pattern_train_result_'+str(isphrase),'r')
pattern_testset=json.load(f)
f.close()
#clf2=SVC()
clf2=LinearSVC()
clf2.fit(pattern_testset[0],pattern_testset[1])
count=0
cnt_var=0
pat_val=-1
bin_val=-1
med_val = median_result('samplereview.txt',isphrase)
temp_class1= clf.predict(proximity_tagger.bin_list('samplereview.txt',isphrase))
if temp_class1 == [1]:
bin_val=1
temp_class2= clf2.predict(proximity_tagger.pattern_list('samplereview.txt',isphrase))
numsum=sum(temp_class2)
if numsum > (len(temp_class2)/2):
pat_val=1
isreviewpositive = 0
if (med_val+bin_val+pat_val)>0:
isreviewpositive = 1
return isreviewpositive
