Datasets:
Tasks:
Question Answering
Modalities:
Text
Formats:
parquet
Languages:
Norwegian Bokmål
Size:
1K - 10K
License:
Update README.md
Browse files
README.md
CHANGED
@@ -47,31 +47,14 @@ NorQuAD is the first Norwegian question answering dataset for machine reading co
|
|
47 |
### Dataset Description
|
48 |
|
49 |
<!-- Provide a longer summary of what this dataset is. -->
|
50 |
-
The dataset
|
51 |
-
pairs for those passages, and (iii) human validation of (a subset of) created question-answer pairs.
|
52 |
-
|
53 |
-
#### Text selection
|
54 |
-
|
55 |
-
Data was selected from openly available sources from Wikipedia and News data.
|
56 |
-
|
57 |
-
|
58 |
-
#### Question-Answer Pairs
|
59 |
|
60 |
|
61 |
|
62 |
|
63 |
-
|
64 |
-
|
65 |
-
In a separate stage, the annotators validated a subset of the NorQuAD dataset. In this phase each
|
66 |
-
annotator replied to the questions created by the
|
67 |
-
other annotator. We chose the question-answer
|
68 |
-
pairs for validation at random. In total, 1378 questions from the set of question-answer pairs, were
|
69 |
-
answered by validators.
|
70 |
-
|
71 |
-
|
72 |
-
- **Curated by:** Annotators (see above).
|
73 |
- **Funded by [optional]:** The UiO Teksthub initiative
|
74 |
-
- **Shared by [optional]:**
|
75 |
- **Language(s) (NLP):** Norwegian Bokmål
|
76 |
- **License:** CC0-1.0
|
77 |
|
@@ -150,9 +133,33 @@ total of 4,752 question-answer pairs.
|
|
150 |
|
151 |
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
|
152 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
The annotation guidelines provided to the annotators are available (here)[https://github.com/ltgoslo/NorQuAD/blob/main/guidelines.md].
|
154 |
For annotation, we used the Haystack annotation tool which was designed for QA collection.
|
155 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
156 |
[More Information Needed]
|
157 |
|
158 |
#### Who are the annotators?
|
|
|
47 |
### Dataset Description
|
48 |
|
49 |
<!-- Provide a longer summary of what this dataset is. -->
|
50 |
+
The dataset provides Norwegian question-answer pairs taken from two data sources: Wikipedia and news.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
|
52 |
|
53 |
|
54 |
|
55 |
+
- **Curated by:** Human annotators (see above).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
- **Funded by [optional]:** The UiO Teksthub initiative
|
57 |
+
- **Shared by [optional]:** The Language Technology Group, University of Oslo
|
58 |
- **Language(s) (NLP):** Norwegian Bokmål
|
59 |
- **License:** CC0-1.0
|
60 |
|
|
|
133 |
|
134 |
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
|
135 |
|
136 |
+
The dataset was created in three stages: (i) selecting text passages, (ii) collecting question-answer
|
137 |
+
pairs for those passages, and (iii) human validation of (a subset of) created question-answer pairs.
|
138 |
+
|
139 |
+
#### Text selection
|
140 |
+
|
141 |
+
Data was selected from openly available sources from Wikipedia and News data, as described above.
|
142 |
+
|
143 |
+
|
144 |
+
#### Question-Answer Pairs
|
145 |
+
|
146 |
+
The annotators were provided with a set of initial instructions, largely based on those for similar datasets, in particular, the English SQuAD
|
147 |
+
dataset (Rajpurkar et al., 2016) and the GermanQuAD data (Moller et al., 2021). These instructions were subsequently refined following regular
|
148 |
+
meetings with the annotation team.
|
149 |
The annotation guidelines provided to the annotators are available (here)[https://github.com/ltgoslo/NorQuAD/blob/main/guidelines.md].
|
150 |
For annotation, we used the Haystack annotation tool which was designed for QA collection.
|
151 |
|
152 |
+
|
153 |
+
#### Human validation
|
154 |
+
|
155 |
+
In a separate stage, the annotators validated a subset of the NorQuAD dataset. In this phase each
|
156 |
+
annotator replied to the questions created by the
|
157 |
+
other annotator. We chose the question-answer
|
158 |
+
pairs for validation at random. In total, 1378 questions from the set of question-answer pairs, were
|
159 |
+
answered by validators.
|
160 |
+
|
161 |
+
|
162 |
+
|
163 |
[More Information Needed]
|
164 |
|
165 |
#### Who are the annotators?
|