Moved to B-I format
Browse files- swe-nerc.py +29 -9
swe-nerc.py
CHANGED
|
@@ -19,15 +19,24 @@
|
|
| 19 |
|
| 20 |
|
| 21 |
LABELS = [
|
| 22 |
-
"EVN",
|
| 23 |
-
"GRO",
|
| 24 |
-
"LOC",
|
| 25 |
-
"MNT",
|
| 26 |
"O",
|
| 27 |
-
"PRS",
|
| 28 |
-
"SMP",
|
| 29 |
-
"TME",
|
| 30 |
-
"WRK"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
]
|
| 32 |
|
| 33 |
|
|
@@ -108,6 +117,7 @@ class SweNERC(datasets.GeneratorBasedBuilder):
|
|
| 108 |
guid = 0
|
| 109 |
tokens = []
|
| 110 |
ner_tags = []
|
|
|
|
| 111 |
for line in f:
|
| 112 |
if line.startswith("-DOCSTART-") or line == "" or line == "\n":
|
| 113 |
if tokens:
|
|
@@ -119,15 +129,25 @@ class SweNERC(datasets.GeneratorBasedBuilder):
|
|
| 119 |
guid += 1
|
| 120 |
tokens = []
|
| 121 |
ner_tags = []
|
|
|
|
| 122 |
else:
|
| 123 |
# tokens are tab separated
|
| 124 |
splits = line.split("\t")
|
| 125 |
tokens.append(splits[0])
|
| 126 |
try:
|
| 127 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 128 |
except:
|
| 129 |
print(splits)
|
| 130 |
raise
|
|
|
|
| 131 |
# last example
|
| 132 |
yield guid, {
|
| 133 |
"id": str(guid),
|
|
|
|
| 19 |
|
| 20 |
|
| 21 |
LABELS = [
|
| 22 |
+
"B-EVN",
|
| 23 |
+
"B-GRO",
|
| 24 |
+
"B-LOC",
|
| 25 |
+
"B-MNT",
|
| 26 |
"O",
|
| 27 |
+
"B-PRS",
|
| 28 |
+
"B-SMP",
|
| 29 |
+
"B-TME",
|
| 30 |
+
"B-WRK",
|
| 31 |
+
"I-EVN",
|
| 32 |
+
"I-GRO",
|
| 33 |
+
"I-LOC",
|
| 34 |
+
"I-MNT",
|
| 35 |
+
"O",
|
| 36 |
+
"I-PRS",
|
| 37 |
+
"I-SMP",
|
| 38 |
+
"I-TME",
|
| 39 |
+
"I-WRK"
|
| 40 |
]
|
| 41 |
|
| 42 |
|
|
|
|
| 117 |
guid = 0
|
| 118 |
tokens = []
|
| 119 |
ner_tags = []
|
| 120 |
+
last_tag = None
|
| 121 |
for line in f:
|
| 122 |
if line.startswith("-DOCSTART-") or line == "" or line == "\n":
|
| 123 |
if tokens:
|
|
|
|
| 129 |
guid += 1
|
| 130 |
tokens = []
|
| 131 |
ner_tags = []
|
| 132 |
+
last_tag = None
|
| 133 |
else:
|
| 134 |
# tokens are tab separated
|
| 135 |
splits = line.split("\t")
|
| 136 |
tokens.append(splits[0])
|
| 137 |
try:
|
| 138 |
+
tag = splits[1].rstrip()
|
| 139 |
+
if tag == "O":
|
| 140 |
+
pass
|
| 141 |
+
elif tag == last_tag:
|
| 142 |
+
tag = "I-" + tag
|
| 143 |
+
else:
|
| 144 |
+
tag = "B-" + tag
|
| 145 |
+
ner_tags.append(tag)
|
| 146 |
+
last_tag = splits[1].rstrip()
|
| 147 |
except:
|
| 148 |
print(splits)
|
| 149 |
raise
|
| 150 |
+
|
| 151 |
# last example
|
| 152 |
yield guid, {
|
| 153 |
"id": str(guid),
|