gabrielaltay
commited on
Commit
•
200f5f2
1
Parent(s):
aae2aaa
initial commit
Browse files- LICENSE +26 -0
- README.md +49 -0
- bigbiohub.py +153 -0
- bioasq_task_b.py +795 -0
LICENSE
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
name: National Library of Medicine Terms and Conditions
|
2 |
+
short_name: NLM_LICENSE
|
3 |
+
|
4 |
+
|
5 |
+
National Library of Medicine Terms and Conditions
|
6 |
+
|
7 |
+
INTRODUCTION
|
8 |
+
|
9 |
+
Downloading data from the National Library of Medicine FTP servers indicates your acceptance of the following Terms and Conditions: No charges, usage fees or royalties are paid to NLM for this data.
|
10 |
+
|
11 |
+
GENERAL TERMS AND CONDITIONS
|
12 |
+
|
13 |
+
Users of the data agree to:
|
14 |
+
acknowledge NLM as the source of the data by including the phrase "Courtesy of the U.S. National Library of Medicine" in a clear and conspicuous manner,
|
15 |
+
properly use registration and/or trademark symbols when referring to NLM products, and
|
16 |
+
not indicate or imply that NLM has endorsed its products/services/applications.
|
17 |
+
|
18 |
+
Users who republish or redistribute the data (services, products or raw data) agree to:
|
19 |
+
maintain the most current version of all distributed data, or
|
20 |
+
make known in a clear and conspicuous manner that the products/services/applications do not reflect the most current/accurate data available from NLM.
|
21 |
+
|
22 |
+
These data are produced with a reasonable standard of care, but NLM makes no warranties express or implied, including no warranty of merchantability or fitness for particular purpose, regarding the accuracy or completeness of the data. Users agree to hold NLM and the U.S. Government harmless from any liability resulting from errors in the data. NLM disclaims any liability for any consequences due to use, misuse, or interpretation of information contained or not contained in the data.
|
23 |
+
|
24 |
+
NLM does not provide legal advice regarding copyright, fair use, or other aspects of intellectual property rights. See the NLM Copyright page.
|
25 |
+
|
26 |
+
NLM reserves the right to change the type and format of its machine-readable data. NLM will take reasonable steps to inform users of any changes to the format of the data before the data are distributed via the announcement section or subscription to email and RSS updates.
|
README.md
CHANGED
@@ -1,3 +1,52 @@
|
|
1 |
---
|
|
|
2 |
license: other
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
language: en
|
3 |
license: other
|
4 |
+
multilinguality: monolingual
|
5 |
+
pretty_name: BioASQ Task B
|
6 |
---
|
7 |
+
|
8 |
+
|
9 |
+
# Dataset Card for BioASQ Task B
|
10 |
+
|
11 |
+
## Dataset Description
|
12 |
+
|
13 |
+
- **Homepage:** http://participants-area.bioasq.org/datasets/
|
14 |
+
- **Pubmed:** True
|
15 |
+
- **Public:** False
|
16 |
+
- **Tasks:** Question Answering
|
17 |
+
|
18 |
+
|
19 |
+
The BioASQ corpus contains multiple question
|
20 |
+
answering tasks annotated by biomedical experts, including yes/no, factoid, list,
|
21 |
+
and summary questions. Pertaining to our objective of comparing neural language
|
22 |
+
models, we focus on the the yes/no questions (Task 7b), and leave the inclusion
|
23 |
+
of other tasks to future work. Each question is paired with a reference text
|
24 |
+
containing multiple sentences from a PubMed abstract and a yes/no answer. We use
|
25 |
+
the official train/dev/test split of 670/75/140 questions.
|
26 |
+
|
27 |
+
See 'Domain-Specific Language Model Pretraining for Biomedical
|
28 |
+
Natural Language Processing'
|
29 |
+
|
30 |
+
|
31 |
+
## Citation Information
|
32 |
+
|
33 |
+
```
|
34 |
+
@article{tsatsaronis2015overview,
|
35 |
+
title = {
|
36 |
+
An overview of the BIOASQ large-scale biomedical semantic indexing and
|
37 |
+
question answering competition
|
38 |
+
},
|
39 |
+
author = {
|
40 |
+
Tsatsaronis, George and Balikas, Georgios and Malakasiotis, Prodromos
|
41 |
+
and Partalas, Ioannis and Zschunke, Matthias and Alvers, Michael R and
|
42 |
+
Weissenborn, Dirk and Krithara, Anastasia and Petridis, Sergios and
|
43 |
+
Polychronopoulos, Dimitris and others
|
44 |
+
},
|
45 |
+
year = 2015,
|
46 |
+
journal = {BMC bioinformatics},
|
47 |
+
publisher = {BioMed Central Ltd},
|
48 |
+
volume = 16,
|
49 |
+
number = 1,
|
50 |
+
pages = 138
|
51 |
+
}
|
52 |
+
```
|
bigbiohub.py
ADDED
@@ -0,0 +1,153 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from dataclasses import dataclass
|
2 |
+
from enum import Enum
|
3 |
+
import datasets
|
4 |
+
from types import SimpleNamespace
|
5 |
+
|
6 |
+
|
7 |
+
BigBioValues = SimpleNamespace(NULL="<BB_NULL_STR>")
|
8 |
+
|
9 |
+
|
10 |
+
@dataclass
|
11 |
+
class BigBioConfig(datasets.BuilderConfig):
|
12 |
+
"""BuilderConfig for BigBio."""
|
13 |
+
|
14 |
+
name: str = None
|
15 |
+
version: datasets.Version = None
|
16 |
+
description: str = None
|
17 |
+
schema: str = None
|
18 |
+
subset_id: str = None
|
19 |
+
|
20 |
+
|
21 |
+
class Tasks(Enum):
|
22 |
+
NAMED_ENTITY_RECOGNITION = "NER"
|
23 |
+
NAMED_ENTITY_DISAMBIGUATION = "NED"
|
24 |
+
EVENT_EXTRACTION = "EE"
|
25 |
+
RELATION_EXTRACTION = "RE"
|
26 |
+
COREFERENCE_RESOLUTION = "COREF"
|
27 |
+
QUESTION_ANSWERING = "QA"
|
28 |
+
TEXTUAL_ENTAILMENT = "TE"
|
29 |
+
SEMANTIC_SIMILARITY = "STS"
|
30 |
+
TEXT_PAIRS_CLASSIFICATION = "TXT2CLASS"
|
31 |
+
PARAPHRASING = "PARA"
|
32 |
+
TRANSLATION = "TRANSL"
|
33 |
+
SUMMARIZATION = "SUM"
|
34 |
+
TEXT_CLASSIFICATION = "TXTCLASS"
|
35 |
+
|
36 |
+
|
37 |
+
entailment_features = datasets.Features(
|
38 |
+
{
|
39 |
+
"id": datasets.Value("string"),
|
40 |
+
"premise": datasets.Value("string"),
|
41 |
+
"hypothesis": datasets.Value("string"),
|
42 |
+
"label": datasets.Value("string"),
|
43 |
+
}
|
44 |
+
)
|
45 |
+
|
46 |
+
pairs_features = datasets.Features(
|
47 |
+
{
|
48 |
+
"id": datasets.Value("string"),
|
49 |
+
"document_id": datasets.Value("string"),
|
50 |
+
"text_1": datasets.Value("string"),
|
51 |
+
"text_2": datasets.Value("string"),
|
52 |
+
"label": datasets.Value("string"),
|
53 |
+
}
|
54 |
+
)
|
55 |
+
|
56 |
+
qa_features = datasets.Features(
|
57 |
+
{
|
58 |
+
"id": datasets.Value("string"),
|
59 |
+
"question_id": datasets.Value("string"),
|
60 |
+
"document_id": datasets.Value("string"),
|
61 |
+
"question": datasets.Value("string"),
|
62 |
+
"type": datasets.Value("string"),
|
63 |
+
"choices": [datasets.Value("string")],
|
64 |
+
"context": datasets.Value("string"),
|
65 |
+
"answer": datasets.Sequence(datasets.Value("string")),
|
66 |
+
}
|
67 |
+
)
|
68 |
+
|
69 |
+
text_features = datasets.Features(
|
70 |
+
{
|
71 |
+
"id": datasets.Value("string"),
|
72 |
+
"document_id": datasets.Value("string"),
|
73 |
+
"text": datasets.Value("string"),
|
74 |
+
"labels": [datasets.Value("string")],
|
75 |
+
}
|
76 |
+
)
|
77 |
+
|
78 |
+
text2text_features = datasets.Features(
|
79 |
+
{
|
80 |
+
"id": datasets.Value("string"),
|
81 |
+
"document_id": datasets.Value("string"),
|
82 |
+
"text_1": datasets.Value("string"),
|
83 |
+
"text_2": datasets.Value("string"),
|
84 |
+
"text_1_name": datasets.Value("string"),
|
85 |
+
"text_2_name": datasets.Value("string"),
|
86 |
+
}
|
87 |
+
)
|
88 |
+
|
89 |
+
kb_features = datasets.Features(
|
90 |
+
{
|
91 |
+
"id": datasets.Value("string"),
|
92 |
+
"document_id": datasets.Value("string"),
|
93 |
+
"passages": [
|
94 |
+
{
|
95 |
+
"id": datasets.Value("string"),
|
96 |
+
"type": datasets.Value("string"),
|
97 |
+
"text": datasets.Sequence(datasets.Value("string")),
|
98 |
+
"offsets": datasets.Sequence([datasets.Value("int32")]),
|
99 |
+
}
|
100 |
+
],
|
101 |
+
"entities": [
|
102 |
+
{
|
103 |
+
"id": datasets.Value("string"),
|
104 |
+
"type": datasets.Value("string"),
|
105 |
+
"text": datasets.Sequence(datasets.Value("string")),
|
106 |
+
"offsets": datasets.Sequence([datasets.Value("int32")]),
|
107 |
+
"normalized": [
|
108 |
+
{
|
109 |
+
"db_name": datasets.Value("string"),
|
110 |
+
"db_id": datasets.Value("string"),
|
111 |
+
}
|
112 |
+
],
|
113 |
+
}
|
114 |
+
],
|
115 |
+
"events": [
|
116 |
+
{
|
117 |
+
"id": datasets.Value("string"),
|
118 |
+
"type": datasets.Value("string"),
|
119 |
+
# refers to the text_bound_annotation of the trigger
|
120 |
+
"trigger": {
|
121 |
+
"text": datasets.Sequence(datasets.Value("string")),
|
122 |
+
"offsets": datasets.Sequence([datasets.Value("int32")]),
|
123 |
+
},
|
124 |
+
"arguments": [
|
125 |
+
{
|
126 |
+
"role": datasets.Value("string"),
|
127 |
+
"ref_id": datasets.Value("string"),
|
128 |
+
}
|
129 |
+
],
|
130 |
+
}
|
131 |
+
],
|
132 |
+
"coreferences": [
|
133 |
+
{
|
134 |
+
"id": datasets.Value("string"),
|
135 |
+
"entity_ids": datasets.Sequence(datasets.Value("string")),
|
136 |
+
}
|
137 |
+
],
|
138 |
+
"relations": [
|
139 |
+
{
|
140 |
+
"id": datasets.Value("string"),
|
141 |
+
"type": datasets.Value("string"),
|
142 |
+
"arg1_id": datasets.Value("string"),
|
143 |
+
"arg2_id": datasets.Value("string"),
|
144 |
+
"normalized": [
|
145 |
+
{
|
146 |
+
"db_name": datasets.Value("string"),
|
147 |
+
"db_id": datasets.Value("string"),
|
148 |
+
}
|
149 |
+
],
|
150 |
+
}
|
151 |
+
],
|
152 |
+
}
|
153 |
+
)
|
bioasq_task_b.py
ADDED
@@ -0,0 +1,795 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2022 The HuggingFace Datasets Authors and the current dataset script contributor.
|
3 |
+
#
|
4 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
+
# you may not use this file except in compliance with the License.
|
6 |
+
# You may obtain a copy of the License at
|
7 |
+
#
|
8 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
9 |
+
#
|
10 |
+
# Unless required by applicable law or agreed to in writing, software
|
11 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
12 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13 |
+
# See the License for the specific language governing permissions and
|
14 |
+
# limitations under the License.
|
15 |
+
"""
|
16 |
+
BioASQ Task B On Biomedical Semantic QA (Involves IR, QA, Summarization qnd
|
17 |
+
More). This task uses benchmark datasets containing development and test
|
18 |
+
questions, in English, along with gold standard (reference) answers constructed
|
19 |
+
by a team of biomedical experts. The participants have to respond with relevant
|
20 |
+
concepts, articles, snippets and RDF triples, from designated resources, as well
|
21 |
+
as exact and 'ideal' answers.
|
22 |
+
|
23 |
+
Fore more information about the challenge, the organisers and the relevant
|
24 |
+
publications please visit: http://bioasq.org/
|
25 |
+
"""
|
26 |
+
import glob
|
27 |
+
import json
|
28 |
+
import os
|
29 |
+
import re
|
30 |
+
|
31 |
+
import datasets
|
32 |
+
|
33 |
+
from .bigbiohub import qa_features
|
34 |
+
from .bigbiohub import BigBioConfig
|
35 |
+
from .bigbiohub import Tasks
|
36 |
+
|
37 |
+
_LANGUAGES = ["English"]
|
38 |
+
_PUBMED = True
|
39 |
+
_LOCAL = True
|
40 |
+
_CITATION = """\
|
41 |
+
@article{tsatsaronis2015overview,
|
42 |
+
title = {
|
43 |
+
An overview of the BIOASQ large-scale biomedical semantic indexing and
|
44 |
+
question answering competition
|
45 |
+
},
|
46 |
+
author = {
|
47 |
+
Tsatsaronis, George and Balikas, Georgios and Malakasiotis, Prodromos
|
48 |
+
and Partalas, Ioannis and Zschunke, Matthias and Alvers, Michael R and
|
49 |
+
Weissenborn, Dirk and Krithara, Anastasia and Petridis, Sergios and
|
50 |
+
Polychronopoulos, Dimitris and others
|
51 |
+
},
|
52 |
+
year = 2015,
|
53 |
+
journal = {BMC bioinformatics},
|
54 |
+
publisher = {BioMed Central Ltd},
|
55 |
+
volume = 16,
|
56 |
+
number = 1,
|
57 |
+
pages = 138
|
58 |
+
}
|
59 |
+
"""
|
60 |
+
|
61 |
+
_DATASETNAME = "bioasq_task_b"
|
62 |
+
_DISPLAYNAME = "BioASQ Task B"
|
63 |
+
|
64 |
+
_BIOASQ_10B_DESCRIPTION = """\
|
65 |
+
The data are intended to be used as training and development data for BioASQ
|
66 |
+
10, which will take place during 2022. There is one file containing the data:
|
67 |
+
- training10b.json
|
68 |
+
|
69 |
+
The file contains the data of the first nine editions of the challenge: 4234
|
70 |
+
questions [1] with their relevant documents, snippets, concepts and RDF
|
71 |
+
triples, exact and ideal answers.
|
72 |
+
|
73 |
+
Differences with BioASQ-training9b.json
|
74 |
+
- 492 new questions added from BioASQ9
|
75 |
+
- The question with id 56c1f01eef6e394741000046 had identical body with
|
76 |
+
602498cb1cb411341a00009e. All relevant elements from both questions
|
77 |
+
are available in the merged question with id 602498cb1cb411341a00009e.
|
78 |
+
- The question with id 5c7039207c78d69471000065 had identical body with
|
79 |
+
601c317a1cb411341a000014. All relevant elements from both questions
|
80 |
+
are available in the merged question with id 601c317a1cb411341a000014.
|
81 |
+
- The question with id 5e4b540b6d0a27794100001c had identical body with
|
82 |
+
602828b11cb411341a0000fc. All relevant elements from both questions
|
83 |
+
are available in the merged question with id 602828b11cb411341a0000fc.
|
84 |
+
- The question with id 5fdb42fba43ad31278000027 had identical body with
|
85 |
+
5d35eb01b3a638076300000f. All relevant elements from both questions
|
86 |
+
are available in the merged question with id 5d35eb01b3a638076300000f.
|
87 |
+
- The question with id 601d76311cb411341a000045 had identical body with
|
88 |
+
6060732b94d57fd87900003d. All relevant elements from both questions
|
89 |
+
are available in the merged question with id 6060732b94d57fd87900003d.
|
90 |
+
|
91 |
+
[1] 4234 questions : 1252 factoid, 1148 yesno, 1018 summary, 816 list
|
92 |
+
"""
|
93 |
+
|
94 |
+
_BIOASQ_9B_DESCRIPTION = """\
|
95 |
+
The data are intended to be used as training and development data for BioASQ 9,
|
96 |
+
which will take place during 2021. There is one file containing the data:
|
97 |
+
- training9b.json
|
98 |
+
|
99 |
+
The file contains the data of the first seven editions of the challenge: 3742
|
100 |
+
questions [1] with their relevant documents, snippets, concepts and RDF triples,
|
101 |
+
exact and ideal answers.
|
102 |
+
|
103 |
+
Differences with BioASQ-training8b.json
|
104 |
+
- 499 new questions added from BioASQ8
|
105 |
+
- The question with id 5e30e689fbd6abf43b00003a had identical body with
|
106 |
+
5880e417713cbdfd3d000001. All relevant elements from both questions
|
107 |
+
are available in the merged question with id 5880e417713cbdfd3d000001.
|
108 |
+
|
109 |
+
[1] 3742 questions : 1091 factoid, 1033 yesno, 899 summary, 719 list
|
110 |
+
"""
|
111 |
+
|
112 |
+
_BIOASQ_8B_DESCRIPTION = """\
|
113 |
+
The data are intended to be used as training and development data for BioASQ 8,
|
114 |
+
which will take place during 2020. There is one file containing the data:
|
115 |
+
- training8b.json
|
116 |
+
|
117 |
+
The file contains the data of the first seven editions of the challenge: 3243
|
118 |
+
questions [1] with their relevant documents, snippets, concepts and RDF triples,
|
119 |
+
exact and ideal answers.
|
120 |
+
|
121 |
+
Differences with BioASQ-training7b.json
|
122 |
+
- 500 new questions added from BioASQ7
|
123 |
+
- 4 questions were removed
|
124 |
+
- The question with id 5717fb557de986d80d000009 had identical body with
|
125 |
+
571e06447de986d80d000016. All relevant elements from both questions
|
126 |
+
are available in the merged question with id 571e06447de986d80d000016.
|
127 |
+
- The question with id 5c589ddb86df2b917400000b had identical body with
|
128 |
+
5c6b7a9e7c78d69471000029. All relevant elements from both questions
|
129 |
+
are available in the merged question with id 5c6b7a9e7c78d69471000029.
|
130 |
+
- The question with id 52ffb5d12059c6d71c00007c had identical body with
|
131 |
+
52e7870a98d023950500001a. All relevant elements from both questions
|
132 |
+
are available in the merged question with id 52e7870a98d023950500001a.
|
133 |
+
- The question with id 53359338d6d3ac6a3400004f had identical body with
|
134 |
+
589a246878275d0c4a000030. All relevant elements from both questions
|
135 |
+
are available in the merged question with id 589a246878275d0c4a000030.
|
136 |
+
|
137 |
+
**** UPDATE 25/02/2020 *****
|
138 |
+
The previous version of the dataset contained an inconsistency on question with
|
139 |
+
id "5c9904eaecadf2e73f00002e", where the "ideal_answer" field was missing.
|
140 |
+
This has been fixed.
|
141 |
+
"""
|
142 |
+
|
143 |
+
_BIOASQ_7B_DESCRIPTION = """\
|
144 |
+
The data are intended to be used as training and development data for BioASQ 7,
|
145 |
+
which will take place during 2019. There is one file containing the data:
|
146 |
+
- BioASQ-trainingDataset7b.json
|
147 |
+
|
148 |
+
The file contains the data of the first six editions of the challenge: 2747
|
149 |
+
questions [1] with their relevant documents, snippets, concepts and RDF triples,
|
150 |
+
exact and ideal answers.
|
151 |
+
|
152 |
+
Differences with BioASQ-trainingDataset6b.json
|
153 |
+
- 500 new questions added from BioASQ6
|
154 |
+
- 4 questions were removed
|
155 |
+
- The question with id 569ed752ceceede94d000004 had identical body with
|
156 |
+
a new question from BioASQ6. All relevant elements from both questions
|
157 |
+
are available in the merged question with id 5abd31e0fcf456587200002c
|
158 |
+
- 3 questions were removed as incomplete: 54d643023706e89528000007,
|
159 |
+
532819afd6d3ac6a3400000f, 517545168ed59a060a00002b
|
160 |
+
- 4 questions were revised for various confusions that have been identified
|
161 |
+
- In 2 questions the ideal answer has been revised :
|
162 |
+
51406e6223fec90375000009, 5172f8118ed59a060a000019
|
163 |
+
- In 4 questions the snippets and documents list has been revised :
|
164 |
+
51406e6223fec90375000009, 5172f8118ed59a060a000019,
|
165 |
+
51593dc8d24251bc05000099, 5158a5b8d24251bc05000097
|
166 |
+
- In 198 questions the documents list has updated with missing
|
167 |
+
documents from the relevant snippets list. [2]
|
168 |
+
|
169 |
+
[1] 2747 questions : 779 factoid, 745 yesno, 667 summary, 556 list
|
170 |
+
[2] 55031181e9bde69634000014, 51406e6223fec90375000009, 54d643023706e89528000007,
|
171 |
+
52bf1b0a03868f1b06000009, 52bf19c503868f1b06000001, 51593dc8d24251bc05000099,
|
172 |
+
530a5117970c65fa6b000007, 553a8d78f321868558000003, 531a3fe3b166e2b806000038,
|
173 |
+
532819afd6d3ac6a3400000f, 5158a5b8d24251bc05000097, 553653a5bc4f83e828000007,
|
174 |
+
535d2cf09a4572de6f000004, 53386282d6d3ac6a3400005a, 517a8ce98ed59a060a000045,
|
175 |
+
55391ce8bc4f83e828000018, 5547d700f35db75526000007, 5713bf261174fb1755000011,
|
176 |
+
6f15c5a2ac5ed1459000012, 52b2e498f828ad283c000010, 570a7594cf1c325851000026,
|
177 |
+
530cefaaad0bf1360c000012, 530f685c329f5fcf1e000002, 550c4011a103b78016000009,
|
178 |
+
552faababc4f83e828000005, 54cf48acf693c3b16b00000b, 550313aae9bde6963400001f,
|
179 |
+
551177626a8cde6b72000005, 54eded8c94afd6150400000c, 550c3754a103b78016000007,
|
180 |
+
56f555b609dd18d46b000007, 54c26e29f693c3b16b000003, 54da0c524b1fd0d33c00000b,
|
181 |
+
52bf1d3c03868f1b0600000d, 5343bdd6aeec6fbd07000001, 52cb9b9b03868f1b0600002d,
|
182 |
+
55423875ec76f5e50c000002, 571366ba1174fb1755000005, 56c4d14ab04e159d0e000003,
|
183 |
+
550c44d1a103b7801600000a, 5547a01cf35db75526000005, 55422640ccca0ce74b000004,
|
184 |
+
54ecb66d445c3b5a5f000002, 553656c4bc4f83e828000009, 5172f8118ed59a060a000019,
|
185 |
+
513711055274a5fb0700000e, 54d892ee014675820d000005, 52e6c92598d0239505000019,
|
186 |
+
5353aedb288f4dae47000006, 52bf1f1303868f1b06000014, 5519113b622b19434500000f,
|
187 |
+
52b2f1724003448f5500000b, 5525317687ecba3764000007, 554a0cadf35db7552600000f,
|
188 |
+
55152bd246478f2f2c000002, 516c3960298dcd4e51000073, 571e417bbb137a4b0c00000a,
|
189 |
+
551910d3622b194345000008, 54dc8ed6c0bb8dce23000002, 511a4ec01159fa8212000004,
|
190 |
+
54d8ea2c4b1fd0d33c000002, 5148e1d6d24251bc0500003a, 515dbb3b298dcd4e51000018,
|
191 |
+
56f7c15a09dd18d46b000012, 51475d5cd24251bc0500001b, 54db7c4ac0bb8dce23000001,
|
192 |
+
57152ebbcb4ef8864c000002, 57134d511174fb1755000002, 55149f156a8cde6b72000013,
|
193 |
+
56bcd422d36b5da378000005, 54ede5c394afd61504000006, 517545168ed59a060a00002b,
|
194 |
+
5710ed19a5ed216440000003, 53442472aeec6fbd07000008, 55088e412e93f0133a000001,
|
195 |
+
54d762653706e89528000014, 550aef0ec2af5d5b7000000a, 552435602c8b63434a000009,
|
196 |
+
552446612c8b63434a00000c, 54d901ec4b1fd0d33c000006, 54cf45e7f693c3b16b00000a,
|
197 |
+
52fc8b772059c6d71c00006e, 5314d05adae131f84700000d, 5512c91b6a8cde6b7200000b,
|
198 |
+
56c5a7605795f9a73e000002, 55030a6ce9bde6963400000f, 553fac39c6a5098552000001,
|
199 |
+
531a3a58b166e2b806000037, 5509bd6a1180f13250000002, 54f9c40ddd3fc62544000001,
|
200 |
+
553c8fd1f32186855800000a, 56bce51cd36b5da37800000a, 550316a6e9bde69634000029,
|
201 |
+
55031286e9bde6963400001b, 536e46f27d100faa09000012, 5502abd1e9bde69634000008,
|
202 |
+
551af9106b348bb82c000002, 54edeb4394afd6150400000b, 5717cdd2070aa3d072000001,
|
203 |
+
56c5ade15795f9a73e000003, 531464a6e3eabad021000014, 58a0d87a78275d0c4a000053,
|
204 |
+
58a3160d60087bc10a00000a, 58a5d54860087bc10a000025, 58a0da5278275d0c4a000054,
|
205 |
+
58a3264e60087bc10a00000d, 589c8ef878275d0c4a000042, 58a3428d60087bc10a00001b,
|
206 |
+
58a3196360087bc10a00000b, 58a341eb60087bc10a000018, 58a3275960087bc10a00000f,
|
207 |
+
58a342e760087bc10a00001c, 58bd645702b8c60953000010, 58bc8e5002b8c60953000006,
|
208 |
+
58bc8e7a02b8c60953000007, 58a1da4e78275d0c4a000059, 58bcb83d02b8c6095300000f,
|
209 |
+
58bc9a5002b8c60953000008, 589dee3778275d0c4a000050, 58a32efe60087bc10a000013,
|
210 |
+
58a327bf60087bc10a000011, 58bca08702b8c6095300000a, 58bc9dbb02b8c60953000009,
|
211 |
+
58c99fcc02b8c60953000029, 58bca2f302b8c6095300000c, 58cbf1f402b8c60953000036,
|
212 |
+
58cdb41302b8c60953000042, 58cdb80302b8c60953000043, 58cdbaf302b8c60953000044,
|
213 |
+
58cb305c02b8c60953000032, 58caf86f02b8c60953000030, 58c1b2f702b8c6095300001e,
|
214 |
+
58bde18b02b8c60953000014, 58eb7898eda5a57672000006, 58caf88c02b8c60953000031,
|
215 |
+
58e11bf76fddd3e83e00000c, 58cdbbd102b8c60953000045, 58df779d6fddd3e83e000001,
|
216 |
+
58dbb4f08acda3452900001a, 58dbb8968acda3452900001b, 58add7699ef3c34033000009,
|
217 |
+
58dbbbf08acda3452900001d, 58dbba438acda3452900001c, 58dd2cb08acda34529000029,
|
218 |
+
58eb9542eda5a57672000007, 58f3ca5c70f9fc6f0f00000d, 58e9e7aa3e8b6dc87c00000d,
|
219 |
+
58e3d9ab3e8b6dc87c000002, 58eb4ce7eda5a57672000004, 58f3c8f470f9fc6f0f00000c,
|
220 |
+
58f3c62970f9fc6f0f00000b, 58adca6d9ef3c34033000007, 58f4b3ee70f9fc6f0f000013,
|
221 |
+
593ff22b70f9fc6f0f000023, 5a679875b750ff4455000004, 5a774585faa1ab7d2e000005,
|
222 |
+
5a6f7245b750ff4455000050, 5a787544faa1ab7d2e00000b, 5a74d9980384be9551000008,
|
223 |
+
5a6a02a3b750ff4455000021, 5a6e47b1b750ff4455000049, 5a87124561bb38fb24000001,
|
224 |
+
5a6e42f1b750ff4455000046, 5a8b1264fcd1d6a10c00001d, 5a981e66fcd1d6a10c00002f,
|
225 |
+
5a8718c861bb38fb24000008, 5a7615af83b0d9ea6600001f, 5a87140a61bb38fb24000003,
|
226 |
+
5a77072c9e632bc06600000a, 5a897601fcd1d6a10c000008, 5a871a6861bb38fb24000009,
|
227 |
+
5a74e9ad0384be955100000a, 5a79d25dfaa1ab7d2e00000f, 5a6900ebb750ff445500001d,
|
228 |
+
5a87145861bb38fb24000004, 5a871b8d61bb38fb2400000a, 5a897a06fcd1d6a10c00000b,
|
229 |
+
5a8dc6b4fcd1d6a10c000026, 5a8712af61bb38fb24000002, 5a8714e261bb38fb24000005,
|
230 |
+
5aa304f1d6d6b54f79000004, 5a981bcffcd1d6a10c00002d, 5aa3fa73d6d6b54f79000008,
|
231 |
+
5aa55b45d6d6b54f7900000d, 5a981dd0fcd1d6a10c00002e, 5a9700adfcd1d6a10c00002c,
|
232 |
+
5a9d8ffe1d1251d03b000022, 5a96c74cfcd1d6a10c000029, 5aa50086d6d6b54f7900000c,
|
233 |
+
5a95765bfcd1d6a10c000028, 5a96f40cfcd1d6a10c00002b, 5ab144fefcf4565872000012,
|
234 |
+
5aa67b4fd6d6b54f7900000f, 5abd5a62fcf4565872000031, 5abbe429fcf456587200001c,
|
235 |
+
5aaef38dfcf456587200000f, 5abce6acfcf4565872000022, 5aae6499fcf456587200000c
|
236 |
+
"""
|
237 |
+
|
238 |
+
_BIOASQ_6B_DESCRIPTION = """\
|
239 |
+
The data are intended to be used as training and development data for BioASQ 6,
|
240 |
+
which will take place during 2018. There is one file containing the data:
|
241 |
+
- BioASQ-trainingDataset6b.json
|
242 |
+
|
243 |
+
Differences with BioASQ-trainingDataset5b.json
|
244 |
+
- 500 new questions added from BioASQ5
|
245 |
+
- 48 pairs of questions with identical bodies have been merged into one
|
246 |
+
question having only one question-id, but all the documents, snippets,
|
247 |
+
concepts, RDF triples and answers of both questions of the pair.
|
248 |
+
- This normalization lead to the removal of 48 deprecated question
|
249 |
+
ids [2] from the dataset and to the update of the 48 remaining
|
250 |
+
questions [3].
|
251 |
+
- In cases where a pair of questions with identical bodies had some
|
252 |
+
inconsistency (e.g. different question type), the inconsistency has
|
253 |
+
been solved merging the pair manually consulting the BioASQ expert team.
|
254 |
+
- 12 questions were revised for various confusions that have been
|
255 |
+
identified
|
256 |
+
- In 8 questions the question type has been changed to better suit to
|
257 |
+
the question body. The change of type lead to corresponding changes
|
258 |
+
in exact answers existence and format : 54fc4e2e6ea36a810c000003,
|
259 |
+
530b01a6970c65fa6b000008, 530cf54dab4de4de0c000009,
|
260 |
+
531b2fc3b166e2b80600003c, 532819afd6d3ac6a3400000f,
|
261 |
+
532aad53d6d3ac6a34000010, 5710ade4cf1c32585100002c,
|
262 |
+
52f65f372059c6d71c000027
|
263 |
+
- In 6 questions the ideal answer has been revised :
|
264 |
+
532aad53d6d3ac6a34000010, 5710ade4cf1c32585100002c,
|
265 |
+
53147b52e3eabad021000015, 5147c8a6d24251bc05000027,
|
266 |
+
5509bd6a1180f13250000002, 58bbb71f22d3005309000016
|
267 |
+
- In 5 questions the exact answer has been revised :
|
268 |
+
5314bd7ddae131f847000006, 53130a77e3eabad02100000f,
|
269 |
+
53148a07dae131f847000002, 53147b52e3eabad021000015,
|
270 |
+
5147c8a6d24251bc05000027
|
271 |
+
- In 2 questions the question body has been revised :
|
272 |
+
52f65f372059c6d71c000027, 5503145ee9bde69634000022
|
273 |
+
- In lists of ideal answers, documents, snippets, concepts and RDF triples
|
274 |
+
any duplicate identical elements have been removed.
|
275 |
+
- Ideal answers in format of one string have been converted to a list with
|
276 |
+
one element for consistency with cases where more than one golden ideal
|
277 |
+
answers are available. (i.e. "ideal_ans1" converted to ["ideal_ans1"])
|
278 |
+
- For yesno questions: All exact answers have been normalized to "yes" or
|
279 |
+
"no" (replacing "Yes", "YES" and "No")
|
280 |
+
- For factoid questions: The format of the exact answer was normalized to a
|
281 |
+
list of strings for each question, representing a set of synonyms
|
282 |
+
answering the question (i.e. [`ans1`, `syn11`, ... ]).
|
283 |
+
- For list questions: The format of the exact answer was normalized to a
|
284 |
+
list of lists. Each internal list represents one element of the answer
|
285 |
+
as a set of synonyms
|
286 |
+
(i.e. [[`ans1`, `syn11`, `syn12`], [`ans2`], [`ans3`, `syn31`] ...]).
|
287 |
+
- Empty elements, e.g. empty lists of documents have been removed.
|
288 |
+
|
289 |
+
[1] 2251 questions : 619 factoid, 616 yesno, 531 summary, 485 list
|
290 |
+
[2] The 48 deprecated question ids are : 52f8b2902059c6d71c000053,
|
291 |
+
52f11bf22059c6d71c000005, 52f77edb2059c6d71c000028, 52ed795098d0239505000032,
|
292 |
+
56d1a9baab2fed4a47000002, 52f7d3472059c6d71c00002f, 52fbe2bf2059c6d71c00006c,
|
293 |
+
52ec961098d023950500002a, 52e8e98298d0239505000020, 56cae5125795f9a73e000024,
|
294 |
+
530cefaaad0bf1360c000007, 530cefaaad0bf1360c000005, 52d63b2803868f1b0600003a,
|
295 |
+
530cefaaad0bf1360c00000a, 516425ff298dcd4e51000051, 55191149622b194345000010,
|
296 |
+
52fa70142059c6d71c000056, 52f77f4d2059c6d71c00002a, 52efc016c8da89891000001a,
|
297 |
+
52efc001c8da898910000019, 52f896ae2059c6d71c000045, 52eceada98d023950500002d,
|
298 |
+
52efc05cc8da89891000001c, 515e078e298dcd4e51000031, 52fe54252059c6d71c000079,
|
299 |
+
514217a6d24251bc05000005, 52d1389303868f1b06000032, 530cf4d5e2bfff940c000003,
|
300 |
+
52fc946d2059c6d71c000071, 52e8e99e98d0239505000021, 52ef7786c8da898910000015,
|
301 |
+
52d8494698d0239505000007, 530cf51d5610acba0c000001, 52f637972059c6d71c000025,
|
302 |
+
52e9f99798d0239505000025, 515de572298dcd4e51000021, 52fe4ad52059c6d71c000077,
|
303 |
+
52f65bf02059c6d71c000026, 52e8e9d298d0239505000022, 52fa74052059c6d71c00005a,
|
304 |
+
52ffbddf2059c6d71c00007d, 56bc932aac7ad1001900001c, 56c02883ef6e394741000017,
|
305 |
+
52d2b75403868f1b06000035, 52f118aa2059c6d71c000003, 52e929eb98d0239505000023,
|
306 |
+
532c12f2d6d3ac6a3400001d, 52d8466298d0239505000006'
|
307 |
+
[3] The 48 questions resulting from merging with their pair have the
|
308 |
+
following ids: 5149aafcd24251bc05000045, 515db020298dcd4e51000011,
|
309 |
+
515db54c298dcd4e51000016, 51680a49298dcd4e51000062, 52b06a68f828ad283c000005,
|
310 |
+
52bf1aa503868f1b06000006, 52bf1af803868f1b06000008, 52bf1d6003868f1b0600000e,
|
311 |
+
52cb9b9b03868f1b0600002d, 52d2818403868f1b06000033, 52df887498d023950500000c,
|
312 |
+
52e0c9a298d0239505000010, 52e203bc98d0239505000011, 52e62bae98d0239505000015,
|
313 |
+
52e6c92598d0239505000019, 52e7bbf698d023950500001d, 52ea605098d0239505000028,
|
314 |
+
52ece29f98d023950500002c, 52ecf2dd98d023950500002e, 52ef7754c8da898910000014,
|
315 |
+
52f112bb2059c6d71c000002, 52f65f372059c6d71c000027, 52f77f752059c6d71c00002b,
|
316 |
+
52f77f892059c6d71c00002c, 52f89ee42059c6d71c00004d, 52f89f4f2059c6d71c00004e,
|
317 |
+
52f89fba2059c6d71c00004f, 52f89fc62059c6d71c000050, 52f89fd32059c6d71c000051,
|
318 |
+
52fa6ac72059c6d71c000055, 52fa73c62059c6d71c000058, 52fa73e82059c6d71c000059,
|
319 |
+
52fa74252059c6d71c00005b, 52fc8b772059c6d71c00006e, 52fc94572059c6d71c000070,
|
320 |
+
52fc94ae2059c6d71c000073, 52fc94db2059c6d71c000074, 52fe52702059c6d71c000078,
|
321 |
+
52fe58f82059c6d71c00007a, 530cefaaad0bf1360c000008, 530cefaaad0bf1360c000010,
|
322 |
+
533ba218fd9a95ea0d000007, 534bb147aeec6fbd07000014, 55167dec46478f2f2c00000a,
|
323 |
+
56c04412ef6e39474100001b, 56c1f01eef6e394741000046, 56c81fd15795f9a73e00000c,
|
324 |
+
587d016ed673c3eb14000002
|
325 |
+
"""
|
326 |
+
|
327 |
+
_BIOASQ_5B_DESCRIPTION = """\
|
328 |
+
The data are intended to be used as training and development data for BioASQ 5,
|
329 |
+
which will take place during 2017. There is one file containing the data:
|
330 |
+
- BioASQ-trainingDataset5b.json
|
331 |
+
|
332 |
+
The file contains the data of the first four editions of the challenge: 1799
|
333 |
+
questions with their relevant documents, snippets, concepts and rdf triples,
|
334 |
+
exact and ideal answers.
|
335 |
+
"""
|
336 |
+
|
337 |
+
_BIOASQ_4B_DESCRIPTION = """\
|
338 |
+
The data are intended to be used as training and development data for BioASQ 4,
|
339 |
+
which will take place during 2016. There is one file containing the data:
|
340 |
+
- BioASQ-trainingDataset4b.json
|
341 |
+
|
342 |
+
The file contains the data of the first three editions of the challenge: 1307
|
343 |
+
questions with their relevant documents, snippets, concepts and rdf triples,
|
344 |
+
exact and ideal answers from the first two editions and 497 questions with
|
345 |
+
similar annotations from the third editions of the challenge.
|
346 |
+
"""
|
347 |
+
|
348 |
+
_BIOASQ_3B_DESCRIPTION = """No README provided."""
|
349 |
+
|
350 |
+
_BIOASQ_2B_DESCRIPTION = """No README provided."""
|
351 |
+
|
352 |
+
_BIOASQ_BLURB_DESCRIPTION = """The BioASQ corpus contains multiple question
|
353 |
+
answering tasks annotated by biomedical experts, including yes/no, factoid, list,
|
354 |
+
and summary questions. Pertaining to our objective of comparing neural language
|
355 |
+
models, we focus on the the yes/no questions (Task 7b), and leave the inclusion
|
356 |
+
of other tasks to future work. Each question is paired with a reference text
|
357 |
+
containing multiple sentences from a PubMed abstract and a yes/no answer. We use
|
358 |
+
the official train/dev/test split of 670/75/140 questions.
|
359 |
+
|
360 |
+
See 'Domain-Specific Language Model Pretraining for Biomedical
|
361 |
+
Natural Language Processing' """
|
362 |
+
|
363 |
+
_DESCRIPTION = {
|
364 |
+
"bioasq_10b": _BIOASQ_10B_DESCRIPTION,
|
365 |
+
"bioasq_9b": _BIOASQ_9B_DESCRIPTION,
|
366 |
+
"bioasq_8b": _BIOASQ_8B_DESCRIPTION,
|
367 |
+
"bioasq_7b": _BIOASQ_7B_DESCRIPTION,
|
368 |
+
"bioasq_6b": _BIOASQ_6B_DESCRIPTION,
|
369 |
+
"bioasq_5b": _BIOASQ_5B_DESCRIPTION,
|
370 |
+
"bioasq_4b": _BIOASQ_4B_DESCRIPTION,
|
371 |
+
"bioasq_3b": _BIOASQ_3B_DESCRIPTION,
|
372 |
+
"bioasq_2b": _BIOASQ_2B_DESCRIPTION,
|
373 |
+
"bioasq_blurb": _BIOASQ_BLURB_DESCRIPTION,
|
374 |
+
}
|
375 |
+
|
376 |
+
_HOMEPAGE = "http://participants-area.bioasq.org/datasets/"
|
377 |
+
|
378 |
+
# Data access reqires registering with BioASQ.
|
379 |
+
# See http://participants-area.bioasq.org/accounts/register/
|
380 |
+
_LICENSE = "NLM_LICENSE"
|
381 |
+
|
382 |
+
_URLs = {
|
383 |
+
"bioasq_10b": ["BioASQ-training10b.zip", None],
|
384 |
+
"bioasq_9b": ["BioASQ-training9b.zip", "Task9BGoldenEnriched.zip"],
|
385 |
+
"bioasq_8b": ["BioASQ-training8b.zip", "Task8BGoldenEnriched.zip"],
|
386 |
+
"bioasq_7b": ["BioASQ-training7b.zip", "Task7BGoldenEnriched.zip"],
|
387 |
+
"bioasq_6b": ["BioASQ-training6b.zip", "Task6BGoldenEnriched.zip"],
|
388 |
+
"bioasq_5b": ["BioASQ-training5b.zip", "Task5BGoldenEnriched.zip"],
|
389 |
+
"bioasq_4b": ["BioASQ-training4b.zip", "Task4BGoldenEnriched.zip"],
|
390 |
+
"bioasq_3b": ["BioASQ-trainingDataset3b.zip", "Task3BGoldenEnriched.zip"],
|
391 |
+
"bioasq_2b": ["BioASQ-trainingDataset2b.zip", "Task2BGoldenEnriched.zip"],
|
392 |
+
"bioasq_blurb": ["BioASQ-training7b.zip", "Task7BGoldenEnriched.zip"],
|
393 |
+
}
|
394 |
+
|
395 |
+
# BLURB train and dev contain all yesno questions from the offical training split
|
396 |
+
# test is all yesno question from the official test split
|
397 |
+
_BLURB_SPLITS = {
|
398 |
+
"dev": {
|
399 |
+
"5313b049e3eabad021000013",
|
400 |
+
"553a8d78f321868558000003",
|
401 |
+
"5158a5b8d24251bc05000097",
|
402 |
+
"571e3d42bb137a4b0c000007",
|
403 |
+
"5175b97a8ed59a060a00002f",
|
404 |
+
"56c9e9d15795f9a73e00001d",
|
405 |
+
"56d19ffaab2fed4a47000001",
|
406 |
+
"518ccac0310faafe0800000b",
|
407 |
+
"56f12ca92ac5ed145900000e",
|
408 |
+
"51680a49298dcd4e51000062",
|
409 |
+
"5339ed7bd6d3ac6a34000060",
|
410 |
+
"516e5f33298dcd4e5100007e",
|
411 |
+
"5327139ad6d3ac6a3400000d",
|
412 |
+
"54e12ae3ae9738404b000004",
|
413 |
+
"5321b8579b2d7acc7e000008",
|
414 |
+
"514a4679d24251bc0500005b",
|
415 |
+
"54c12fd1f693c3b16b000001",
|
416 |
+
"52df887498d023950500000c",
|
417 |
+
"52f20d802059c6d71c00000a",
|
418 |
+
"532f0c4ed6d3ac6a3400002e",
|
419 |
+
"52b2f3b74003448f5500000c",
|
420 |
+
"52b2f1724003448f5500000b",
|
421 |
+
"515d9a42298dcd4e5100000d",
|
422 |
+
"5159b990d24251bc050000a3",
|
423 |
+
"54e12c30ae9738404b000005",
|
424 |
+
"553a6a9fbc4f83e82800001c",
|
425 |
+
"5509ec41c2af5d5b70000006",
|
426 |
+
"56cae40b5795f9a73e000022",
|
427 |
+
"51680b0e298dcd4e51000065",
|
428 |
+
"515df89e298dcd4e5100002f",
|
429 |
+
"54f49e56d0d681a040000004",
|
430 |
+
"571e3e2abb137a4b0c000008",
|
431 |
+
"515debe7298dcd4e51000026",
|
432 |
+
"56f6ab7009dd18d46b00000d",
|
433 |
+
"53302bced6d3ac6a34000039",
|
434 |
+
"5322de919b2d7acc7e000012",
|
435 |
+
"5709f212cf1c325851000020",
|
436 |
+
"5502abd1e9bde69634000008",
|
437 |
+
"516c220e298dcd4e51000071",
|
438 |
+
"5894597e7d9090f353000004",
|
439 |
+
"5895ec5e7d9090f353000015",
|
440 |
+
"58bbb8ae22d3005309000018",
|
441 |
+
"58bc58c302b8c60953000001",
|
442 |
+
"58c276bc02b8c60953000020",
|
443 |
+
"58c0825502b8c6095300001b",
|
444 |
+
"58ab1f6c9ef3c34033000002",
|
445 |
+
"58adbe999ef3c34033000005",
|
446 |
+
"58df3e408acda3452900002d",
|
447 |
+
"58dfec676fddd3e83e000006",
|
448 |
+
"58d8d0cc8acda34529000008",
|
449 |
+
"58b67fae22d3005309000009",
|
450 |
+
"58dbbbf08acda3452900001d",
|
451 |
+
"58dbba438acda3452900001c",
|
452 |
+
"58dbbdac8acda3452900001e",
|
453 |
+
"58dcbb8c8acda34529000021",
|
454 |
+
"5a468785966455904c00000d",
|
455 |
+
"5a70de5199e2c3af26000005",
|
456 |
+
"5a67a550b750ff4455000009",
|
457 |
+
"5a679875b750ff4455000004",
|
458 |
+
"5a7a44b4faa1ab7d2e000010",
|
459 |
+
"5a67ade5b750ff445500000c",
|
460 |
+
"5a8881118cb19eca6b000006",
|
461 |
+
"5a67b48cb750ff4455000010",
|
462 |
+
"5a679be1b750ff4455000005",
|
463 |
+
"5a7340962dc08e987e000017",
|
464 |
+
"5a737e233b9d13c70800000d",
|
465 |
+
"5a8dc57ffcd1d6a10c000025",
|
466 |
+
"5a6d186db750ff4455000031",
|
467 |
+
"5a70d43b99e2c3af26000003",
|
468 |
+
"5a70ec6899e2c3af2600000c",
|
469 |
+
"5a9ac4161d1251d03b000010",
|
470 |
+
"5a733d2a2dc08e987e000015",
|
471 |
+
"5a74acd80384be9551000006",
|
472 |
+
"5aa6800ad6d6b54f79000011",
|
473 |
+
"5a9d9ab94e03427e73000003",
|
474 |
+
}
|
475 |
+
}
|
476 |
+
|
477 |
+
_SUPPORTED_TASKS = [Tasks.QUESTION_ANSWERING]
|
478 |
+
_SOURCE_VERSION = "1.0.0"
|
479 |
+
_BIGBIO_VERSION = "1.0.0"
|
480 |
+
|
481 |
+
|
482 |
+
class BioasqTaskBDataset(datasets.GeneratorBasedBuilder):
|
483 |
+
"""
|
484 |
+
BioASQ Task B On Biomedical Semantic QA.
|
485 |
+
Creates configs for BioASQ2 through BioASQ10.
|
486 |
+
"""
|
487 |
+
|
488 |
+
DEFAULT_CONFIG_NAME = "bioasq_9b_source"
|
489 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
490 |
+
BIGBIO_VERSION = datasets.Version(_BIGBIO_VERSION)
|
491 |
+
|
492 |
+
# BioASQ2 through BioASQ10
|
493 |
+
BUILDER_CONFIGS = []
|
494 |
+
for version in range(2, 11):
|
495 |
+
BUILDER_CONFIGS.append(
|
496 |
+
BigBioConfig(
|
497 |
+
name=f"bioasq_{version}b_source",
|
498 |
+
version=SOURCE_VERSION,
|
499 |
+
description=f"bioasq{version} Task B source schema",
|
500 |
+
schema="source",
|
501 |
+
subset_id=f"bioasq_{version}b",
|
502 |
+
)
|
503 |
+
)
|
504 |
+
|
505 |
+
BUILDER_CONFIGS.append(
|
506 |
+
BigBioConfig(
|
507 |
+
name=f"bioasq_{version}b_bigbio_qa",
|
508 |
+
version=BIGBIO_VERSION,
|
509 |
+
description=f"bioasq{version} Task B in simplified BigBio schema",
|
510 |
+
schema="bigbio_qa",
|
511 |
+
subset_id=f"bioasq_{version}b",
|
512 |
+
)
|
513 |
+
)
|
514 |
+
|
515 |
+
# BLURB Benchmark config https://microsoft.github.io/BLURB/
|
516 |
+
BUILDER_CONFIGS.append(
|
517 |
+
BigBioConfig(
|
518 |
+
name=f"bioasq_blurb_bigbio_qa",
|
519 |
+
version=BIGBIO_VERSION,
|
520 |
+
description=f"BLURB benchmark in simplified BigBio schema",
|
521 |
+
schema="bigbio_qa",
|
522 |
+
subset_id=f"bioasq_blurb",
|
523 |
+
)
|
524 |
+
)
|
525 |
+
|
526 |
+
def _info(self):
|
527 |
+
|
528 |
+
# BioASQ Task B source schema
|
529 |
+
if self.config.schema == "source":
|
530 |
+
features = datasets.Features(
|
531 |
+
{
|
532 |
+
"id": datasets.Value("string"),
|
533 |
+
"type": datasets.Value("string"),
|
534 |
+
"body": datasets.Value("string"),
|
535 |
+
"documents": datasets.Sequence(datasets.Value("string")),
|
536 |
+
"concepts": datasets.Sequence(datasets.Value("string")),
|
537 |
+
"ideal_answer": datasets.Sequence(datasets.Value("string")),
|
538 |
+
"exact_answer": datasets.Sequence(datasets.Value("string")),
|
539 |
+
"triples": [
|
540 |
+
{
|
541 |
+
"p": datasets.Value("string"),
|
542 |
+
"s": datasets.Value("string"),
|
543 |
+
"o": datasets.Value("string"),
|
544 |
+
}
|
545 |
+
],
|
546 |
+
"snippets": [
|
547 |
+
{
|
548 |
+
"offsetInBeginSection": datasets.Value("int32"),
|
549 |
+
"offsetInEndSection": datasets.Value("int32"),
|
550 |
+
"text": datasets.Value("string"),
|
551 |
+
"beginSection": datasets.Value("string"),
|
552 |
+
"endSection": datasets.Value("string"),
|
553 |
+
"document": datasets.Value("string"),
|
554 |
+
}
|
555 |
+
],
|
556 |
+
}
|
557 |
+
)
|
558 |
+
# simplified schema for QA tasks
|
559 |
+
elif self.config.schema == "bigbio_qa":
|
560 |
+
features = qa_features
|
561 |
+
|
562 |
+
return datasets.DatasetInfo(
|
563 |
+
description=_DESCRIPTION[self.config.subset_id],
|
564 |
+
features=features,
|
565 |
+
supervised_keys=None,
|
566 |
+
homepage=_HOMEPAGE,
|
567 |
+
license=str(_LICENSE),
|
568 |
+
citation=_CITATION,
|
569 |
+
)
|
570 |
+
|
571 |
+
def _dump_gold_json(self, data_dir):
|
572 |
+
"""
|
573 |
+
BioASQ test data is split into multiple records {9B1_golden.json,...,9B5_golden.json}
|
574 |
+
We combine these files into a single test set file 9Bx_golden.json
|
575 |
+
"""
|
576 |
+
# BLURB is based on version 7
|
577 |
+
version = (
|
578 |
+
re.search(r"bioasq_([0-9]+)b", self.config.subset_id).group(1)
|
579 |
+
if "blurb" not in self.config.name
|
580 |
+
else "7"
|
581 |
+
)
|
582 |
+
gold_fpath = os.path.join(
|
583 |
+
data_dir, f"Task{version}BGoldenEnriched/bx_golden.json"
|
584 |
+
)
|
585 |
+
|
586 |
+
if not os.path.exists(gold_fpath):
|
587 |
+
# combine all gold json files
|
588 |
+
filelist = glob.glob(os.path.join(data_dir, "*/*.json"))
|
589 |
+
data = {"questions": []}
|
590 |
+
for fname in sorted(filelist):
|
591 |
+
with open(fname, "rt", encoding="utf-8") as file:
|
592 |
+
data["questions"].extend(json.load(file)["questions"])
|
593 |
+
# dump gold to json
|
594 |
+
with open(gold_fpath, "wt", encoding="utf-8") as file:
|
595 |
+
json.dump(data, file, indent=2)
|
596 |
+
|
597 |
+
return f"Task{version}BGoldenEnriched/bx_golden.json"
|
598 |
+
|
599 |
+
def _blurb_split_generator(self, train_dir, test_dir):
|
600 |
+
"""
|
601 |
+
Create splits for BLURB Benchmark
|
602 |
+
"""
|
603 |
+
gold_fpath = self._dump_gold_json(test_dir)
|
604 |
+
|
605 |
+
# create train/dev splits from yesno questions
|
606 |
+
train_fpath = os.path.join(train_dir, "blurb_bioasq_train.json")
|
607 |
+
dev_fpath = os.path.join(train_dir, "blurb_bioasq_dev.json")
|
608 |
+
|
609 |
+
blurb_splits = {
|
610 |
+
"train": {"questions": []},
|
611 |
+
"dev": {"questions": []},
|
612 |
+
"test": {"questions": []},
|
613 |
+
}
|
614 |
+
|
615 |
+
if not os.path.exists(train_fpath):
|
616 |
+
data_fpath = os.path.join(train_dir, "BioASQ-training7b/trainining7b.json")
|
617 |
+
with open(data_fpath, "rt", encoding="utf-8") as file:
|
618 |
+
data = json.load(file)
|
619 |
+
|
620 |
+
for record in data["questions"]:
|
621 |
+
if record["type"] != "yesno":
|
622 |
+
continue
|
623 |
+
if record["id"] in _BLURB_SPLITS["dev"]:
|
624 |
+
blurb_splits["dev"]["questions"].append(record)
|
625 |
+
else:
|
626 |
+
blurb_splits["train"]["questions"].append(record)
|
627 |
+
|
628 |
+
with open(train_fpath, "wt", encoding="utf-8") as file:
|
629 |
+
json.dump(blurb_splits["train"], file, indent=2)
|
630 |
+
|
631 |
+
with open(dev_fpath, "wt", encoding="utf-8") as file:
|
632 |
+
json.dump(blurb_splits["dev"], file, indent=2)
|
633 |
+
|
634 |
+
# create test split from yesno questions
|
635 |
+
with open(os.path.join(test_dir, gold_fpath), "rt", encoding="utf-8") as file:
|
636 |
+
data = json.load(file)
|
637 |
+
|
638 |
+
for record in data["questions"]:
|
639 |
+
if record["type"] != "yesno":
|
640 |
+
continue
|
641 |
+
blurb_splits["test"]["questions"].append(record)
|
642 |
+
|
643 |
+
test_fpath = os.path.join(test_dir, "blurb_bioasq_test.json")
|
644 |
+
with open(test_fpath, "wt", encoding="utf-8") as file:
|
645 |
+
json.dump(blurb_splits["test"], file, indent=2)
|
646 |
+
|
647 |
+
return [
|
648 |
+
datasets.SplitGenerator(
|
649 |
+
name=datasets.Split.TRAIN,
|
650 |
+
gen_kwargs={
|
651 |
+
"filepath": train_fpath,
|
652 |
+
"split": "train",
|
653 |
+
},
|
654 |
+
),
|
655 |
+
datasets.SplitGenerator(
|
656 |
+
name=datasets.Split.VALIDATION,
|
657 |
+
gen_kwargs={
|
658 |
+
"filepath": dev_fpath,
|
659 |
+
"split": "dev",
|
660 |
+
},
|
661 |
+
),
|
662 |
+
datasets.SplitGenerator(
|
663 |
+
name=datasets.Split.TEST,
|
664 |
+
gen_kwargs={
|
665 |
+
"filepath": test_fpath,
|
666 |
+
"split": "test",
|
667 |
+
},
|
668 |
+
),
|
669 |
+
]
|
670 |
+
|
671 |
+
def _split_generators(self, dl_manager):
|
672 |
+
"""Returns SplitGenerators."""
|
673 |
+
|
674 |
+
if self.config.data_dir is None:
|
675 |
+
raise ValueError(
|
676 |
+
"This is a local dataset. Please pass the data_dir kwarg to load_dataset."
|
677 |
+
)
|
678 |
+
|
679 |
+
train_dir, test_dir = dl_manager.download_and_extract(
|
680 |
+
[
|
681 |
+
os.path.join(self.config.data_dir, _url)
|
682 |
+
for _url in _URLs[self.config.subset_id]
|
683 |
+
]
|
684 |
+
)
|
685 |
+
# create gold dump and get path
|
686 |
+
gold_fpath = self._dump_gold_json(test_dir)
|
687 |
+
|
688 |
+
# older versions of bioasq have different folder formats
|
689 |
+
train_fpaths = {
|
690 |
+
"bioasq_2b": "BioASQ_2013_TaskB/BioASQ-trainingDataset2b.json",
|
691 |
+
"bioasq_3b": "BioASQ-trainingDataset3b.json",
|
692 |
+
"bioasq_4b": "BioASQ-training4b/BioASQ-trainingDataset4b.json",
|
693 |
+
"bioasq_5b": "BioASQ-training5b/BioASQ-trainingDataset5b.json",
|
694 |
+
"bioasq_6b": "BioASQ-training6b/BioASQ-trainingDataset6b.json",
|
695 |
+
"bioasq_7b": "BioASQ-training7b/trainining7b.json",
|
696 |
+
"bioasq_8b": "training8b.json", # HACK - this zipfile strips the dirname
|
697 |
+
"bioasq_9b": "BioASQ-training9b/training9b.json",
|
698 |
+
"bioasq_10b": "BioASQ-training10b/training10b.json",
|
699 |
+
}
|
700 |
+
|
701 |
+
# BLURB has custom train/dev/test splits based on Task 7B
|
702 |
+
if "blurb" in self.config.name:
|
703 |
+
return self._blurb_split_generator(train_dir, test_dir)
|
704 |
+
|
705 |
+
return [
|
706 |
+
datasets.SplitGenerator(
|
707 |
+
name=datasets.Split.TRAIN,
|
708 |
+
gen_kwargs={
|
709 |
+
"filepath": os.path.join(
|
710 |
+
train_dir, train_fpaths[self.config.subset_id]
|
711 |
+
),
|
712 |
+
"split": "train",
|
713 |
+
},
|
714 |
+
),
|
715 |
+
datasets.SplitGenerator(
|
716 |
+
name=datasets.Split.TEST,
|
717 |
+
gen_kwargs={
|
718 |
+
"filepath": os.path.join(test_dir, gold_fpath),
|
719 |
+
"split": "test",
|
720 |
+
},
|
721 |
+
),
|
722 |
+
]
|
723 |
+
|
724 |
+
def _get_exact_answer(self, record):
|
725 |
+
"""The value exact_answer can be in different formats based on question type."""
|
726 |
+
if record["type"] == "yesno":
|
727 |
+
exact_answer = [record["exact_answer"]]
|
728 |
+
elif record["type"] == "summary":
|
729 |
+
exact_answer = []
|
730 |
+
# summary question types only have an ideal answer, so use that for bigbio
|
731 |
+
if self.config.schema == "bigbio_qa":
|
732 |
+
exact_answer = (
|
733 |
+
record["ideal_answer"]
|
734 |
+
if isinstance(record["ideal_answer"], list)
|
735 |
+
else [record["ideal_answer"]]
|
736 |
+
)
|
737 |
+
|
738 |
+
elif record["type"] == "list":
|
739 |
+
exact_answer = record["exact_answer"]
|
740 |
+
elif record["type"] == "factoid":
|
741 |
+
# older version of bioasq sometimes represent this as as string
|
742 |
+
exact_answer = (
|
743 |
+
record["exact_answer"]
|
744 |
+
if isinstance(record["exact_answer"], list)
|
745 |
+
else [record["exact_answer"]]
|
746 |
+
)
|
747 |
+
return exact_answer
|
748 |
+
|
749 |
+
def _generate_examples(self, filepath, split):
|
750 |
+
"""Yields examples as (key, example) tuples."""
|
751 |
+
|
752 |
+
if self.config.schema == "source":
|
753 |
+
with open(filepath, encoding="utf-8") as file:
|
754 |
+
data = json.load(file)
|
755 |
+
for i, record in enumerate(data["questions"]):
|
756 |
+
yield i, {
|
757 |
+
"id": record["id"],
|
758 |
+
"type": record["type"],
|
759 |
+
"body": record["body"],
|
760 |
+
"documents": record["documents"],
|
761 |
+
"concepts": record["concepts"] if "concepts" in record else [],
|
762 |
+
"triples": record["triples"] if "triples" in record else [],
|
763 |
+
"ideal_answer": record["ideal_answer"]
|
764 |
+
if isinstance(record["ideal_answer"], list)
|
765 |
+
else [record["ideal_answer"]],
|
766 |
+
"exact_answer": self._get_exact_answer(record),
|
767 |
+
"snippets": record["snippets"] if "snippets" in record else [],
|
768 |
+
}
|
769 |
+
|
770 |
+
elif self.config.schema == "bigbio_qa":
|
771 |
+
# NOTE: Years 2014-2016 (BioASQ2-BioASQ4) have duplicate records
|
772 |
+
cache = set()
|
773 |
+
with open(filepath, encoding="utf-8") as file:
|
774 |
+
uid = 0
|
775 |
+
data = json.load(file)
|
776 |
+
for record in data["questions"]:
|
777 |
+
# for questions that do not have snippets, skip
|
778 |
+
if "snippets" not in record:
|
779 |
+
continue
|
780 |
+
for i, snippet in enumerate(record["snippets"]):
|
781 |
+
key = f'{record["id"]}_{i}'
|
782 |
+
# ignore duplicate records
|
783 |
+
if key not in cache:
|
784 |
+
cache.add(key)
|
785 |
+
yield uid, {
|
786 |
+
"id": key,
|
787 |
+
"document_id": snippet["document"],
|
788 |
+
"question_id": record["id"],
|
789 |
+
"question": record["body"],
|
790 |
+
"type": record["type"],
|
791 |
+
"choices": [],
|
792 |
+
"context": snippet["text"],
|
793 |
+
"answer": self._get_exact_answer(record),
|
794 |
+
}
|
795 |
+
uid += 1
|