-
Notifications
You must be signed in to change notification settings - Fork 0
/
data_process.py
58 lines (51 loc) · 1.84 KB
/
data_process.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
def build_corpus(split, make_vocab=True):
data_path = 'data/conll03_BIO/' + split + '.txt'
# 读取数据
words_list = []
tags_list = []
with open(data_path, 'r', encoding='utf-8') as f:
words = []
tags = []
for line in f:
if line != '\n': # 列表内嵌列表实现多个句子在一个列表中
word, tag = line.strip('\n').split() # 把换行符去掉
words.append(word)
tags.append(tag)
else:
words_list.append(words)
tags_list.append(tags)
words = [] # 清空列表
tags = []
if make_vocab: # 在训练集上建立映射
word2id = build_map(words_list, False)
tag2id = build_map(tags_list, True)
return words_list, tags_list, word2id, tag2id
else:
return words_list, tags_list
def build_map(lists, tag):
maps = {}
if tag: # 标记是否是标签,因为0映射的数据不同,此处提前规定好0标签的对应值方便之后在lstm补全时用
maps['O'] = 0 # 把O放在第一位上
else:
maps['<pad>'] = 0 # 把pad放在第一位上
for list_ in lists:
for e in list_:
if e not in maps:
maps[e] = len(maps)
return maps
def word_tag_id(word_list, tag_list, word_map, tag_map):
# 把str型转换成id
words_data = []
tags_data = []
for words, tags in zip(word_list, tag_list):
word_data = []
tag_data = []
for word, tag in zip(words, tags):
try:
word_data.append(word_map[word])
except:
word_data.append(word_map['<unk>'])
tag_data.append(tag_map[tag])
words_data.append(word_data)
tags_data.append(tag_data)
return words_data, tags_data