Datasets:
Tasks:
Summarization
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
---
|
|
|
|
| 2 |
language:
|
| 3 |
- ar
|
| 4 |
- zh
|
|
@@ -10,46 +11,21 @@ language:
|
|
| 10 |
- hi
|
| 11 |
- ro
|
| 12 |
- es
|
| 13 |
-
|
| 14 |
-
|
| 15 |
tags:
|
| 16 |
-
- summarisation
|
| 17 |
- multi-document
|
| 18 |
-
-
|
| 19 |
- abstractive
|
| 20 |
-
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
- 1K<n<10K
|
| 27 |
-
dataset_info:
|
| 28 |
-
features:
|
| 29 |
-
- name: language
|
| 30 |
-
dtype: string
|
| 31 |
-
- name: doc_id
|
| 32 |
-
dtype: string
|
| 33 |
-
- name: documents_text
|
| 34 |
-
dtype: string
|
| 35 |
-
- name: summary
|
| 36 |
-
dtype: string
|
| 37 |
-
- name: summary_1
|
| 38 |
-
dtype: string
|
| 39 |
-
- name: summary_2
|
| 40 |
-
dtype: string
|
| 41 |
-
- name: summary_3
|
| 42 |
-
dtype: string
|
| 43 |
-
- name: doc_ids
|
| 44 |
-
dtype: string
|
| 45 |
-
splits:
|
| 46 |
-
- train
|
| 47 |
-
- validation
|
| 48 |
-
- test
|
| 49 |
-
task_templates:
|
| 50 |
-
- text2text-generation
|
| 51 |
---
|
| 52 |
|
|
|
|
| 53 |
# MultiLing Multilingual Summarisation Corpus
|
| 54 |
|
| 55 |
The **MultiLing Multilingual Summarisation Corpus** is a comprehensive multilingual benchmark for **single-document** and **multi-document abstractive summarisation**, originally created for the *MultiLing 2011* and *MultiLing 2013* shared tasks held under ACL.
|
|
|
|
| 1 |
---
|
| 2 |
+
license: "cc-by-4.0"
|
| 3 |
language:
|
| 4 |
- ar
|
| 5 |
- zh
|
|
|
|
| 11 |
- hi
|
| 12 |
- ro
|
| 13 |
- es
|
| 14 |
+
task_categories:
|
| 15 |
+
- summarization
|
| 16 |
tags:
|
|
|
|
| 17 |
- multi-document
|
| 18 |
+
- single-document
|
| 19 |
- abstractive
|
| 20 |
+
- multilingual
|
| 21 |
+
configs:
|
| 22 |
+
- config_name: single
|
| 23 |
+
description: "Single-document summarisation pairs (article + summary)."
|
| 24 |
+
- config_name: multi
|
| 25 |
+
description: "Multi-document summarisation clusters with 3 human summaries."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
---
|
| 27 |
|
| 28 |
+
|
| 29 |
# MultiLing Multilingual Summarisation Corpus
|
| 30 |
|
| 31 |
The **MultiLing Multilingual Summarisation Corpus** is a comprehensive multilingual benchmark for **single-document** and **multi-document abstractive summarisation**, originally created for the *MultiLing 2011* and *MultiLing 2013* shared tasks held under ACL.
|