def update_tag_scheme(sentences, tag_scheme):
"""
更新为制定编码 BIO->BIOES
:param sentences:
:param tag_scheme:
:return:
"""
# 首先是检测编码是否为BIO
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
# 首先是检测编码是否为BIO
if not data_utils.check_bio(tags):
s_str = "\n".join(" ".join(w) for w in s) # "\n".join(" ".join(w) for w in s)
raise Exception("输入的句子应为BIO编码，请检查输入句子%i:\n%s" % (i, s_str))
if tag_scheme == "BIO":
for word, new_tag in zip(s, tags):
word[-1] = new_tag
if tag_scheme == "BIOES":
new_tags = data_utils.bio_to_bioes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception("非法目标编码！既不是BIO也不是BIOES")
def update_tag_scheme(sentences, tag_scheme):
"""
更新为指定编码
:param sentences:
:param tag_scheme:
:return:
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
if not du.check_bio(tags):
s_str = '\n'.join(" ".join(w) for w in s)
raise Exception("输入的句子应为BIO编码，请检查输入句子%i:\n%s" % (i, s_str))
if tag_scheme == 'BIOES':
new_tags = du.bio_to_bioes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
def update_tag_scheme(sentences, tag_scheme):
"""
更新为指定标签
:param sentences:
:param tag_scheme:
:return:
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
if not data_utils.check_bio(tags):
raise Exception('输入的句子应为BIO标注法，请检查%i句' % i)
if tag_scheme == 'BIO':
for word, new_tag in zip(s, tags):
word[-1] = new_tag
if tag_scheme == 'BIOES':
new_tags = data_utils.bio_to_bioes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('非法编码')
def update_tag_scheme(sentences, tag_scheme):
"""
将BIO编码更新为BIOES编码
:param sentences:
:param tag_scheme:
:return:
"""
for i, s in enumerate(sentences):
tags = [w[-1] for w in s]
if not data_utils.check_bio(tags):
s_str = '\n'.join(' '.join(w) for w in s)
raise Exception('输入的句子应为BIO编码, 请检查句子%i:\n%s' % (i, s_str))
if tag_scheme == 'BIO':
for word, tag in zip(s, tags):
word[-1] = tag
elif tag_scheme == 'BIOES':
new_tags = data_utils.bio_to_bioes(tags)
for word, new_tag in zip(s, new_tags):
word[-1] = new_tag
else:
raise Exception('非法编码')