Commit
·
bc92390
1
Parent(s):
2aa3c1d
Update README.md
Browse files
README.md
CHANGED
@@ -73,7 +73,7 @@ LR-Sum is released under a Creative Commons license (CC BY 4.0), making it one o
|
|
73 |
|
74 |
- **Curated by:** BLT Lab: Chester Palen-Michel and Constantine Lignos
|
75 |
- **Shared by:** Chester Palen-Michel
|
76 |
-
- **Language(s) (NLP)
|
77 |
- **License:** CC-BY 4.0
|
78 |
|
79 |
### Dataset Sources [optional]
|
@@ -81,8 +81,9 @@ LR-Sum is released under a Creative Commons license (CC BY 4.0), making it one o
|
|
81 |
Multilingual Open Text v1.6
|
82 |
which is a collection of newswire text from Voice of America (VOA).
|
83 |
|
84 |
-
- **
|
85 |
-
- **
|
|
|
86 |
|
87 |
## Uses
|
88 |
|
@@ -99,71 +100,64 @@ This dataset only includes newswire text, so models trained on the data may not
|
|
99 |
|
100 |
|
101 |
## Dataset Structure
|
|
|
102 |
|
103 |
-
|
104 |
-
|
105 |
-
|
|
|
|
|
106 |
|
107 |
## Dataset Creation
|
108 |
|
109 |
### Curation Rationale
|
110 |
|
111 |
-
|
112 |
-
|
113 |
-
[More Information Needed]
|
114 |
|
115 |
### Source Data
|
116 |
|
117 |
-
|
118 |
|
119 |
#### Data Collection and Processing
|
120 |
|
121 |
-
|
122 |
|
123 |
-
[More Information Needed]
|
124 |
|
125 |
#### Who are the source data producers?
|
126 |
|
127 |
-
|
128 |
-
|
129 |
-
[More Information Needed]
|
130 |
-
|
131 |
-
### Annotations [optional]
|
132 |
-
|
133 |
-
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
|
134 |
|
135 |
#### Annotation process
|
136 |
|
137 |
-
|
138 |
-
|
139 |
-
[More Information Needed]
|
140 |
|
141 |
#### Who are the annotators?
|
142 |
|
143 |
-
|
144 |
|
145 |
-
[More Information Needed]
|
146 |
|
147 |
#### Personal and Sensitive Information
|
148 |
|
149 |
-
|
150 |
-
|
151 |
-
[More Information Needed]
|
152 |
|
153 |
## Bias, Risks, and Limitations
|
154 |
|
155 |
-
|
|
|
156 |
|
157 |
-
[More Information Needed]
|
158 |
|
159 |
### Recommendations
|
160 |
|
161 |
-
|
162 |
-
|
163 |
-
|
|
|
164 |
|
165 |
## Citation
|
166 |
|
|
|
|
|
167 |
**BibTeX:**
|
168 |
```
|
169 |
@inproceedings{palen-michel-lignos-2023-lr,
|
@@ -185,14 +179,10 @@ Users should be made aware of the risks, biases and limitations of the dataset.
|
|
185 |
}
|
186 |
```
|
187 |
|
188 |
-
|
189 |
-
|
190 |
-
Palen-Michel, C. & Lignos, C. (2023). LR-Sum: Summarization for Less-Resourced Languages. In Findings of the Association for Computational Linguistics: ACL 2023, pages 6829–6844, Toronto, Canada. Association for Computational Linguistics.
|
191 |
-
|
192 |
-
## Dataset Card Authors [optional]
|
193 |
|
194 |
-
Chester Palen-Michel
|
195 |
|
196 |
## Dataset Card Contact
|
197 |
|
198 |
-
Chester Palen-Michel
|
|
|
73 |
|
74 |
- **Curated by:** BLT Lab: Chester Palen-Michel and Constantine Lignos
|
75 |
- **Shared by:** Chester Palen-Michel
|
76 |
+
- **Language(s) (NLP):** Albanian, Amharic, Armenian, Azerbaijani, Bengali, Bosnian, Burmese, Chinese, English, French, Georgian, Greek, Haitian Creole, Hausa, Indonesian, Khmer, Kinyarwanda, Korean, Kurdish, Lao, Macedonian, Northern Ndebele, Pashto, Persian, Portuguese, Russian, Serbian, Shona, Somali, Spanish, Swahili, Thai, Tibetan, Tigrinya, Turkish, Ukrainian, Urdu, Uzbek, Vietnamese
|
77 |
- **License:** CC-BY 4.0
|
78 |
|
79 |
### Dataset Sources [optional]
|
|
|
81 |
Multilingual Open Text v1.6
|
82 |
which is a collection of newswire text from Voice of America (VOA).
|
83 |
|
84 |
+
- **Paper:** [https://aclanthology.org/2023.findings-acl.427/](https://aclanthology.org/2023.findings-acl.427/)
|
85 |
+
- **Repository:** [https://github.com/bltlab/lr-sum](https://github.com/bltlab/lr-sum)
|
86 |
+
|
87 |
|
88 |
## Uses
|
89 |
|
|
|
100 |
|
101 |
|
102 |
## Dataset Structure
|
103 |
+
Each field is a string:
|
104 |
|
105 |
+
'id': Article unique id
|
106 |
+
'url': URL for the news article
|
107 |
+
'title': The title of the news article
|
108 |
+
'summary': The summary of the article
|
109 |
+
'text': The full text of the news article not including title
|
110 |
|
111 |
## Dataset Creation
|
112 |
|
113 |
### Curation Rationale
|
114 |
|
115 |
+
Research in automatic summarization for less resourced languages.
|
|
|
|
|
116 |
|
117 |
### Source Data
|
118 |
|
119 |
+
Voice of America (VOA)
|
120 |
|
121 |
#### Data Collection and Processing
|
122 |
|
123 |
+
See our [paper](https://aclanthology.org/2023.findings-acl.427/) for details on collection and processing.
|
124 |
|
|
|
125 |
|
126 |
#### Who are the source data producers?
|
127 |
|
128 |
+
Voice of America (VOA)
|
|
|
|
|
|
|
|
|
|
|
|
|
129 |
|
130 |
#### Annotation process
|
131 |
|
132 |
+
The summaries are found in news article meta data. More detail about the curation process can be found in our paper.
|
|
|
|
|
133 |
|
134 |
#### Who are the annotators?
|
135 |
|
136 |
+
The summaries are found in the news article meta data. The authors of the summaries are authors and staff for VOA.
|
137 |
|
|
|
138 |
|
139 |
#### Personal and Sensitive Information
|
140 |
|
141 |
+
The only sensative personal information would be information already published in news articles on VOA.
|
142 |
+
See [VOA's mission and values](https://www.insidevoa.com/p/5831.html#:~:text=VOA%20has%20a%20legal%20obligation,sites%20at%20the%20earliest%20opportunity.)
|
|
|
143 |
|
144 |
## Bias, Risks, and Limitations
|
145 |
|
146 |
+
The content in this dataset is newswire.
|
147 |
+
See [VOA's mission and values](https://www.insidevoa.com/p/5831.html#:~:text=VOA%20has%20a%20legal%20obligation,sites%20at%20the%20earliest%20opportunity.) for more detail about the journalistic integrity and policy.
|
148 |
|
|
|
149 |
|
150 |
### Recommendations
|
151 |
|
152 |
+
The data is newswire text. Training text generation models on this dataset will have similar risks and limitations to other text generation models including hallucinations and potentially inaccurate statements.
|
153 |
+
For some languages that have fewer examples, issues with text generation models are likely to be more pronounced.
|
154 |
+
The dataset is primarily released for research despite having a permissive license.
|
155 |
+
We encourage users to thoroughly test and evaluate any models trained using this data before putting them into production environments.
|
156 |
|
157 |
## Citation
|
158 |
|
159 |
+
If you make use of this dataset, please cite our paper using this bibtex:
|
160 |
+
|
161 |
**BibTeX:**
|
162 |
```
|
163 |
@inproceedings{palen-michel-lignos-2023-lr,
|
|
|
179 |
}
|
180 |
```
|
181 |
|
182 |
+
## Dataset Card Authors
|
|
|
|
|
|
|
|
|
183 |
|
184 |
+
Chester Palen-Michel [@cpalenmichel](https://github.com/cpalenmichel)
|
185 |
|
186 |
## Dataset Card Contact
|
187 |
|
188 |
+
Chester Palen-Michel [@cpalenmichel](https://github.com/cpalenmichel)
|