def update_tag_scheme(sentences, tag_scheme):
"""
对标注模式进行检查和转换
:param sentences:
:param tag_scheme:
:return:
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# 对IOB标注模式的数据进行处理，确定原数据的标注模式是IOB2格式
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
# 若要转换成IOBES格式
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
# enumerate() 函数用于将一个可遍历的数据对象组合为一个索引序列，同时列出数据和数据下标，一般用在 for 循环当中。
# 这里i为从0开始的序列，i为每个句子的序号
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# 这里 tags 是取sentences中每个字的标签-- w[-1]的意思是取数组最后一位，并将标签按句子排成一个列表
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
# join() 方法用于将序列中的元素以指定的字符<这里为‘ ’和‘\n’>连接生成一个新的字符串。
# 以上将在错误信息中打印
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags) # iob_iobes是根据iob标签形成iobes
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag # 将原标签word[-1]换成新标签
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
序列标记常见标记方案：
http://nathanlvzs.github.io/Several-Tagging-Schemes-for-Sequential-Tagging.html
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
1：检查样本的标签是否为正确的IOB格式，如果不对则纠正。
2：将IOB格式转化为IOBES格式。
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
if not iob(tags):
s_str = '\n'.join(' '.join(w) for w in s)
print('Sentences should be given in IOB format! \n' +
'Please check sentence %i:\n%s' % (i, s_str))
""" 如果用IOB格式训练，则检查并纠正一遍 """
if tag_scheme == 'iob':
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
""" 将IOB格式转化为IOBES格式 """
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
for i, s in enumerate(
sentences): # s是sentence，[['Russian', 'B-MISC'], ['military', 'O']
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(
s, new_tags): # zip将s和new tags两个list合并为一个元组的形式
# word: ['BRUSSELS', 'B-LOC']
# new_tag是转换为iobes后的纯标签
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
# print("tag_schema:{}".format(tag_scheme))
# tag_schema: iobes
for i, s in enumerate(sentences):
# print("i:{}\ts:{}".format(i,s))
tags = [w[-1] for w in s]
# print("tags:{}".format(tags))
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
# print("tags:{}".format(tags))
new_tags = iob_iobes(tags)
# print("nwe tags:{}".format(new_tags))
for word, new_tag in zip(s, new_tags):
# print("word:{}".format(word))
word[-1] = new_tag
# exit()
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
#��鲢����IOB2�ľ��ӱ�ע������������IOB1��IOB2���� ������õ���IOB2��IOBS��IOBES��
for i, s in enumerate(sentences): #sһ������
tags = [w[-1] for w in s] #w[-1]ÿ���ֵı�ע
# Check that tags are given in the IOB format
if not iob2(
tags
): #iob2(tags) ������ÿ���ֵı�ע���ж�(���� IOB1 to IOB2)�Ƿ�һ���Ϸ��ı�ע
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags) #iob ת���� iobes
for word, new_tag in zip(s, new_tags): #new_tags ת����ı���
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format!' +
'please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknow tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
# 该模型的输入是加载好的数据sentences是load_sentences的返回值和tag_scheme是'iob'或'iobes'。简而言之就是增加语句的end和
# 该模型的输入数据是:
# [[['入', 'O'], ['院', 'O'], ['情', 'B-DRU'], ['况', 'I-DRU'], ['：', 'O'], ['女', 'O'], ['，', 'O'], ['0', 'O'],
# ['0', 'O'], ['岁', 'O'], ['，', 'O'], ['以', 'O'], ['突', 'B-SYM'], ['发', 'I-SYM'], ['言', 'B-SYM'], ['语', 'I-SYM'],
# ['不', 'I-SYM'], ['清', 'I-SYM'], ['0', 'O'], ['天', 'O'], ['，', 'O'], ['加', 'O'], ['重', 'O'], ['0', 'O'],
# ['天', 'O'], ['入', 'O'], ['院', 'O'], ['。', 'O'], ['入', 'O'], ['院', 'O'], ['情', 'B-DRU'], ['况', 'I-DRU'],
# ['：', 'O'], ['患', 'O'], ['者', 'O'], ['以', 'O'], ['腰', 'O'], ['痛', 'O'], ['伴', 'O'], ['双', 'B-REG'],
# ['下', 'I-REG'], ['肢', 'I-REG'], ['疼', 'B-SYM'], ['痛', 'I-SYM'], ['半', 'O'], ['年', 'O'], ['，', 'O'], ['加', 'O'],
# ['重', 'O'], ['0', 'O'], ['0', 'O'], ['余', 'O'], ['天', 'O'], ['为', 'O'], ['主', 'O'], ['诉', 'O'], ['入', 'O'],
# ['院', 'O'], ['。', 'O']]]
#
# 该模型的返回值是经过修改的传入的sentences,该处牵扯python的语法，具体内容见https://blog.csdn.net/nathan_yo/article/details/98639051
# 该模型更改后的数据为：
# [[['入', 'O'], ['院', 'O'], ['情', 'B-DRU'], ['况', 'E-DRU'], ['：', 'O'], ['女', 'O'], ['，', 'O'], ['0', 'O'],
# ['0', 'O'], ['岁', 'O'], ['，', 'O'], ['以', 'O'], ['突', 'B-SYM'], ['发', 'E-SYM'], ['言', 'B-SYM'], ['语', 'I-SYM'],
# ['不', 'I-SYM'], ['清', 'E-SYM'], ['0', 'O'], ['天', 'O'], ['，', 'O'], ['加', 'O'], ['重', 'O'], ['0', 'O'],
# ['天', 'O'], ['入', 'O'], ['院', 'O'], ['。', 'O'], ['入', 'O'], ['院', 'O'], ['情', 'B-DRU'], ['况', 'E-DRU'],
# ['：', 'O'], ['患', 'O'], ['者', 'O'], ['以', 'O'], ['腰', 'O'], ['痛', 'O'], ['伴', 'O'], ['双', 'B-REG'],
# ['下', 'I-REG'], ['肢', 'E-REG'], ['疼', 'B-SYM'], ['痛', 'E-SYM'], ['半', 'O'], ['年', 'O'], ['，', 'O'], ['加', 'O'],
# ['重', 'O'], ['0', 'O'], ['0', 'O'], ['余', 'O'], ['天', 'O'], ['为', 'O'], ['主', 'O'], ['诉', 'O'], ['入', 'O'],
# ['院', 'O'], ['。', 'O']]]
# 这里i 和 s分别取字符和标记
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
# 该处用于检查错误，是否全是IOB标签B-
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted.
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')
def update_tag_scheme(sentences, tag_scheme):
"""
Check and update sentences tagging scheme to IOB2.
Only IOB1 and IOB2 schemes are accepted. #为什么只有这两个体系可用？？？
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# Check that tags are given in the IOB format
if not iob2(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('Sentences should be given in IOB format! ' +
'Please check sentence %i:\n%s' % (i, s_str))
#以下转标记体系没看懂？？？
if tag_scheme == 'iob':
# If format was IOB1, we convert to IOB2
for word, new_tag in zip(s, tags):
word[-1] = new_tag
elif tag_scheme == 'iobes':
new_tags = iob_iobes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('Unknown tagging scheme!')