-
Notifications
You must be signed in to change notification settings - Fork 15
/
NamedEntityRecognition_py3.py
294 lines (239 loc) · 9.99 KB
/
NamedEntityRecognition_py3.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
__author__ = 'Tian Kang'
#============ Parser step 1: Entity&Attribute recognition =====#
# #
# sys.argv 1: input dir (input) #
# sys.argv 2: input text name (input) #
# sys.argv 3: output NER xml dir #
#
# email: tk2624@cumc.columbia.edu (Tian) #
# June, 2016 #
# #
#==================================================================#
# for python 3 and metampa 2016 version
import os.path
_PATH = os.path.join( *os.path.split(__file__)[:-1] )
import sys,string,os,re,csv
import codecs
from text_processing import txtconll as t2c
from text_processing import preprocess
from features_dir import POS,BrownClustering,umls_identify
from text_processing import label_from_annotation as labeling
import nltk
#import screen
from nltk.stem.lancaster import LancasterStemmer
from bin import readfromdir
from bin.negex import *
from xml.etree import ElementTree as ET
def txt2matrix_fortest(sent,crf_matrix,filename):
sent=sent.rstrip()
metamap_output=umls_identify.formating_for_metamap(curpath,sent,filename)
one_sent_term,type_list=umls_identify.label_umls_cui(metamap_output,sent)
pos_list=POS.pos_tagging(one_sent_term)
pos_list.append(".")
type_list.append("O")
terms=sent.split()
term_id=0
for t in one_sent_term:
term=t
lemma=st.stem(term)
#vector=word2vec.ouput_embedding(model,term.lower(),50)
bc=BrownClustering.bc_indexing(term.lower(),bc_index)
#print >>crf_matrix, t+"\t"+lemma+"\t"+pos_list[term_id]+"\t"+type_list[term_id]+"\t"+bc
crf_matrix.write(t+"\t"+lemma+"\t"+pos_list[term_id]+"\t"+type_list[term_id]+"\t"+bc+"\n")
term_id+=1
#print >>crf_matrix
crf_matrix.write("\n")
def txt2matrix_fortrain(ann_dir,mytrain,tag_included,filename,curpath):
txt_files=readfromdir.get_file_list(ann_dir,['txt'])
print ("there's "+ str(len(txt_files))+" in total!")
i=0
for txt_file in txt_files:
i+=1
# read files
myraw=codecs.open(txt_file).read()
match=re.search('^(.*)\.txt',txt_file)
name=match.group(1)
#ann_file=name+'_new.ann'
ann_file=name+'.ann'
print ("reading file from",txt_file,ann_file,"...")
myann=codecs.open(ann_file,"r")
# output features
text_tagged=labeling.ann_tagging(myann,myraw,tag_included)
lines=" ".join(text_tagged.split(r'[;\n]'))
sents=nltk.sent_tokenize(lines)
lines=" ### ".join(sents)
term_list, tag_list,index_list=t2c.txt2conll(lines,1) # "1" here represents it's a training texts with annoatioin; "0" represents raw texts
sents=" ".join(term_list).split("###")
type_list=[]
pos_list=[]
# extract umls concepts:
j=0
for sent in sents:
if j>=len(term_list):
break
one_sent_list = sent.split()
metamap_output=umls_identify.formating_for_metamap(curpath,sent,filename)
one_sent_term,type_list=umls_identify.label_umls_cui(metamap_output,sent)
pos_list=POS.pos_tagging(one_sent_term)
#pos_list= POS.pos_tagging(one_sent_list)
#pos_list.append(".")
#type_list.append("O")
terms=sent.split()
sent_id=0
for t in terms:
if term_list[j]== "###":
j=j+1
term=term_list[j]
lemma=st.stem(term)
#vector=word2vec.ouput_embedding(model,term.lower(),50)
#bc=BrownClustering.bc_indexing(term.lower(),bc_index)
mytrain.write(
term_list[j] + "\t" + lemma + "\t" + pos_list[sent_id] + "\t" + type_list[sent_id] + "\t" +
tag_list[j] + "\n")
#mytrain.write(term_list[j]+"\t"+lemma+"\t"+pos_list[sent_id]+"\t"+type_list[sent_id]+"\t"+tag_list[j]+"\n")
#mytrain.write(term_list[j] + "\t" + lemma + "\t" + pos_list[sent_id] + "\t" +bc+"\t"+ tag_list[j] + "\n")
sent_id+=1
j=j+1
mytrain.write("\n")
if i%5==0:
print (str(i) +" files finished")
def generate_XML(crfresult_input,NERxml_output):
sents,entity=t2c.conll2txt(crfresult_input)
entity_lists=['Condition','Observation','Drug','Procedure_Device']
attribute_lists=['Qualifier','Measurement','Temporal_measurement']
#print >>NERxml_output,"<?xml version=\"1.0\"?>"
#print >>NERxml_output,"<root>"
NERxml_output.write("<?xml version=\"1.0\"?>\n")
NERxml_output.write("<root>\n")
j=0
for sent in sents:
if sent == "":
continue
clean_sent=t2c.clean_txt(sent)
# clean_sent = re.sub('>=', " largerequalthan ", clean_sent.decode('utf-8'))
# clean_sent = re.sub('<=', " smallerequalthan ", clean_sent)
#print sent
#print "===",entity[j]
pattern='class=\'(\w+)\''
entities=entity[j].split('\n\t\t')
new_entities=[]
for e in entities:
if e =='':
new_entities.append('\n')
continue
match=re.search(pattern,e)
if match.group(1) in attribute_lists:
p1='\<entity'
p2='entity\>'
new=re.sub(p1,'<attribute',e)
new=re.sub(p2,'attribute>',new)
new_entities.append(new)
else:
new_entities.append(e)
entity[j]="\n\t\t".join(new_entities)
#clean_sent=re.sub("\'"," POSSESS ",clean_sent)
#print >>NERxml_output,"\t"+"<sent>\n"+"\t\t<text>"+clean_sent+"</text>"
#print >>NERxml_output,"\t\t"+entity[j]
#print >>NERxml_output,"\t"+"</sent>"
NERxml_output.write("\t"+"<sent>\n"+"\t\t<text>"+clean_sent+"</text>\n")
NERxml_output.write("\t\t"+entity[j]+"\n")
NERxml_output.write("\t"+"</sent>\n")
j+=1
#print >>NERxml_output,"</root>"
NERxml_output.write("</root>\n")
def run_crf( model_dir, matrix_dir, output_dir):
command='crf_test -m '+model_dir+' '+matrix_dir+' > '+output_dir
os.system(command)
def detect_negation(concept,sent,irules):
pattern="^\s?(\w?.*\w?)\s?"
match=re.search(pattern,concept)
clean_concept=match.group(1)
words=re.split("\s+",clean_concept)
# print concept,len(words)
if len(words)>2:
words=[words[-2],words[-1]]
concept=" ".join(words)
# print concept
tagger = negTagger(sentence = sent, phrases =[concept], rules = irules, negP=False)
tag=tagger.getNegationFlag()
negation="N"
if tag=="negated":
negation="Y"
return negation
# load models
st = LancasterStemmer()
bc_index=BrownClustering.read_bcindex("trained_models/brownclustering.index")
entity_lists=['Condition','Observation','Drug','Procedure_Device']
attribute_lists=['Qualifier','Measurement','Temporal_measurement']
PICO_lists = ['Intervention_Primary','Participant_Primary','Outcome_Primary','Intervention_detail',
'Participant_detail','Outcome_detail']
tag_included=entity_lists
tag_included.append('Negation_cue')
curpath = os.path.abspath(os.curdir)
def main():
#===== train =====
# read file:
annotation_dir='/Users/tian/Documents/EliIE-master/Tempfile/annotation/train'
mytrain=codecs.open('/Users/tian/Documents/EliIE-master/Tempfile/pico_train.conll','w')
txt2matrix_fortrain(annotation_dir,mytrain,PICO_lists,'picotrain',curpath)
print ("matrix_finished!")
#myxml=codecs.open('/Users/kangtian/Documents/NER_data/Negation.xml','w')
#myconll=codecs.open('/Users/kangtian/Documents/NER_data/Negation.matrix')
#generate_XML(myconll,myxml)
#print "negation xml finished!"
'''
#========== predict ==========
# read files:
input_dir=sys.argv[1]+"/"+sys.argv[2]
print ("Reading text from ",input_dir)
mytrial=codecs.open(input_dir).read()
match=re.search('^(.*)\.txt',sys.argv[2])
filename=sys.argv[2]
if match:
filename=match.group(1)
nerxmlname=filename+"_NER_temp.xml"
output_dir=sys.argv[3]+'/'+nerxmlname
myxml=codecs.open(output_dir,'w')
matrix_dir='Tempfile/test_'+filename+ '.matrix'
mymatrix=codecs.open(matrix_dir,'w')
crf_result_dir='Tempfile/test_'+filename+'_crf.result'
ori_sents=mytrial.split("\n")
sents=[]
for sent in ori_sents:
s=nltk.sent_tokenize(sent)
sents.extend(s)
# make conll matrix
for sent in sents:
cleansent=preprocess.preprocess(sent)
filteredsent=preprocess.ec_filtering(cleansent)
if filteredsent:
txt2matrix_fortest(filteredsent,mymatrix,filename)
# run crf to predict and generate temp xml file
run_crf('trained_models/bc_umls_pos_lemma_bow.model', matrix_dir ,crf_result_dir)
myconll=codecs.open(crf_result_dir,"r")
generate_XML(myconll,myxml)
# final step: predict negation for entities
NER_tree=ET.ElementTree(file=output_dir)
root = NER_tree.getroot()
rfile = open(r'bin/EC_triggers.txt')
irules = sortRules(rfile.readlines())
for child in root:
sent=''
for child2 in child.findall('text'):
sent=child2.text
for child2 in child.findall('entity'):
if child2.attrib['class']=='Negation_cue':
continue
child2.attrib['negated']="N"
concept=child2.text
neg_tag=detect_negation(concept,sent,irules)
child2.attrib['negated']=neg_tag
print ("negation finished!")
new_tree_name=filename+"_NER.xml"
new_output_dir=sys.argv[3]+'/'+new_tree_name
NER_addneg_tree=codecs.open(new_output_dir,'w')
NER_tree.write(NER_addneg_tree)
rm_comand='rm '+matrix_dir+' '+crf_result_dir+' '+output_dir
os.system(rm_comand)
'''
if __name__ == '__main__': main()