Datasets:
Tasks:
Text Classification
Modalities:
Text
Sub-tasks:
multi-class-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Adding medical subdomain classification
Browse files- data/MedicalDomain/test.csv +0 -0
- data/MedicalDomain/train.csv +0 -0
- raft.py +24 -3
data/MedicalDomain/test.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|
data/MedicalDomain/train.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|
raft.py
CHANGED
@@ -55,6 +55,10 @@ _URLs = {
|
|
55 |
'AIInitiatives': {
|
56 |
'train': "./data/AIInitiatives/train.csv",
|
57 |
'test': "./data/AIInitiatives/test.csv"
|
|
|
|
|
|
|
|
|
58 |
}
|
59 |
} # TODO: Generate these automatically.
|
60 |
|
@@ -83,10 +87,13 @@ class Raft(datasets.GeneratorBasedBuilder):
|
|
83 |
"safety or technical safety."),
|
84 |
datasets.BuilderConfig(name="AIInitiatives-multilabel", version=VERSION,
|
85 |
description="For each initiative, decide which (if any) of Ethics, "
|
86 |
-
"Governance, and Social Good apply to the initiative's AI goals")
|
|
|
|
|
|
|
87 |
]
|
88 |
|
89 |
-
DEFAULT_CONFIG_NAME = "TAISafety
|
90 |
|
91 |
def _info(self):
|
92 |
# TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
|
@@ -117,6 +124,13 @@ class Raft(datasets.GeneratorBasedBuilder):
|
|
117 |
"answer_socialgood": datasets.Value("string"),
|
118 |
}
|
119 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
120 |
return datasets.DatasetInfo(
|
121 |
# This is the description that will appear on the datasets page.
|
122 |
description=_DESCRIPTION,
|
@@ -185,7 +199,7 @@ class Raft(datasets.GeneratorBasedBuilder):
|
|
185 |
stage, date, country, notes, answer_ethics, \
|
186 |
answer_governance, answer_socialgood = row
|
187 |
if split == "test":
|
188 |
-
|
189 |
yield id_, {"name": name,
|
190 |
"organization": organization,
|
191 |
"description": description,
|
@@ -199,3 +213,10 @@ class Raft(datasets.GeneratorBasedBuilder):
|
|
199 |
"answer_ethics": answer_ethics,
|
200 |
"answer_governance": answer_governance,
|
201 |
"answer_socialgood": answer_socialgood}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
55 |
'AIInitiatives': {
|
56 |
'train': "./data/AIInitiatives/train.csv",
|
57 |
'test': "./data/AIInitiatives/test.csv"
|
58 |
+
},
|
59 |
+
'MedicalDomain': {
|
60 |
+
'train': "./data/MedicalDomain/train.csv",
|
61 |
+
'test': "./data/MedicalDomain/test.csv"
|
62 |
}
|
63 |
} # TODO: Generate these automatically.
|
64 |
|
|
|
87 |
"safety or technical safety."),
|
88 |
datasets.BuilderConfig(name="AIInitiatives-multilabel", version=VERSION,
|
89 |
description="For each initiative, decide which (if any) of Ethics, "
|
90 |
+
"Governance, and Social Good apply to the initiative's AI goals"),
|
91 |
+
datasets.BuilderConfig(name="MedicalDomain-multiclass", version=VERSION,
|
92 |
+
description="Which medical subdomain does a given physician's clinical notes "
|
93 |
+
"belong to?"),
|
94 |
]
|
95 |
|
96 |
+
DEFAULT_CONFIG_NAME = "TAISafety-binary" # It's not mandatory to have a default configuration. Just use one if it make sense.
|
97 |
|
98 |
def _info(self):
|
99 |
# TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
|
|
|
124 |
"answer_socialgood": datasets.Value("string"),
|
125 |
}
|
126 |
)
|
127 |
+
elif self.config.name.startswith("MedicalDomain"):
|
128 |
+
features = datasets.Features(
|
129 |
+
{
|
130 |
+
"text": datasets.Value("string"),
|
131 |
+
"answer": datasets.Value("string"),
|
132 |
+
}
|
133 |
+
)
|
134 |
return datasets.DatasetInfo(
|
135 |
# This is the description that will appear on the datasets page.
|
136 |
description=_DESCRIPTION,
|
|
|
199 |
stage, date, country, notes, answer_ethics, \
|
200 |
answer_governance, answer_socialgood = row
|
201 |
if split == "test":
|
202 |
+
answer_ethics, answer_governance, answer_socialgood = "", "", ""
|
203 |
yield id_, {"name": name,
|
204 |
"organization": organization,
|
205 |
"description": description,
|
|
|
213 |
"answer_ethics": answer_ethics,
|
214 |
"answer_governance": answer_governance,
|
215 |
"answer_socialgood": answer_socialgood}
|
216 |
+
if dataset == "MedicalDomain":
|
217 |
+
text, answer = row
|
218 |
+
if split == "test":
|
219 |
+
answer = ""
|
220 |
+
yield id_, {"text": text,
|
221 |
+
"answer": answer}
|
222 |
+
|