Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -129,4 +129,73 @@ configs:
|
|
129 |
path: papyrus-m/test-*
|
130 |
- split: validation
|
131 |
path: papyrus-m/validation-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
132 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
129 |
path: papyrus-m/test-*
|
130 |
- split: validation
|
131 |
path: papyrus-m/validation-*
|
132 |
+
license: apache-2.0
|
133 |
+
language:
|
134 |
+
- en
|
135 |
+
- fr
|
136 |
+
tags:
|
137 |
+
- text-to-text
|
138 |
+
- keyphrase-generation
|
139 |
+
pretty_name: Papyrus
|
140 |
+
size_categories:
|
141 |
+
- 10K<n<100K
|
142 |
---
|
143 |
+
|
144 |
+
|
145 |
+
# Dataset Card for Papyrus
|
146 |
+
|
147 |
+
- **Paper:** [A new dataset for multilingual keyphrase generation](https://proceedings.neurips.cc/paper_files/paper/2022/hash/f88709551258331f9ab31b33c71021a4-Abstract-Datasets_and_Benchmarks.html)
|
148 |
+
- **Github:** <https://github.com/smolPixel/French-keyphrase-generation>
|
149 |
+
|
150 |
+
|
151 |
+
## Dataset Description
|
152 |
+
|
153 |
+
### Dataset Summary
|
154 |
+
|
155 |
+
The datasets are derived from Papyrus, a repository at Université de Montréal containing various types of documents, mainly theses with abstracts in multiple languages, primarily French and English. The entries are provided in four different configurations based on the languages of abstracts, allowing for generating keyphrases in French, English, or multiple languages.
|
156 |
+
- **Papyrus-f:** From the French abstracts, generate French keyphrases.
|
157 |
+
- **Papyrus-e:** From the English abstracts, generate English keyphrases.
|
158 |
+
- **Papyrus-m:** From one abstract in any language, generate keyphrases in that same
|
159 |
+
language (one language to one language).
|
160 |
+
- **Papyrus-a:** From the multiple abstracts of a document, generate keyphrases in the
|
161 |
+
same languages as the abstracts (many to many languages).
|
162 |
+
|
163 |
+
### Languages
|
164 |
+
|
165 |
+
- **Main languages:** English, French
|
166 |
+
- **Others:** Spanish, German, Italian, Portuguese, Arabic, Tagalog, Catalan, Greek, Turkish, Russian, Polish, Farsi, Indonesian, Lingala, Swedish, Finnish, Romanian, Korean
|
167 |
+
|
168 |
+
|
169 |
+
## Dataset Structure
|
170 |
+
|
171 |
+
### Dataset content
|
172 |
+
|
173 |
+
| Config | Train set size | Valid. set size | Test set size |
|
174 |
+
| --------- | -------------- | --------------- | ------------- |
|
175 |
+
| papyrus-m | 20963 | 3040 | 6061 |
|
176 |
+
| papyrus-e | 10508 | 1539 | 3046 |
|
177 |
+
| papyrus-f | 10299 | 1488 | 2981 |
|
178 |
+
| papyrus-a | 11290 | 1638 | 3261 |
|
179 |
+
|
180 |
+
### Data fields
|
181 |
+
|
182 |
+
- **doc_id:** a unique id for the original document.
|
183 |
+
- **title:** title of the thesis or article (the language of the title does not always match the language of the abstract/keyphrases).
|
184 |
+
- **input_text:** abstract of the document.
|
185 |
+
- **keyphrases:** associated keyphrases.
|
186 |
+
- **lang:** language of the abstract/keyphrases.
|
187 |
+
|
188 |
+
|
189 |
+
## Citation
|
190 |
+
|
191 |
+
@inproceedings{NEURIPS2022_f8870955,
|
192 |
+
author = {Piedboeuf, Fr\'{e}d\'{e}ric and Langlais, Philippe},
|
193 |
+
booktitle = {Advances in Neural Information Processing Systems},
|
194 |
+
editor = {S. Koyejo and S. Mohamed and A. Agarwal and D. Belgrave and K. Cho and A. Oh},
|
195 |
+
pages = {38046--38059},
|
196 |
+
publisher = {Curran Associates, Inc.},
|
197 |
+
title = {A new dataset for multilingual keyphrase generation},
|
198 |
+
url = {https://proceedings.neurips.cc/paper_files/paper/2022/file/f88709551258331f9ab31b33c71021a4-Paper-Datasets_and_Benchmarks.pdf},
|
199 |
+
volume = {35},
|
200 |
+
year = {2022}
|
201 |
+
}
|