metadata
task_categories:
- sequence-modeling
task_ids:
- language-modeling
multilinguality:
- translation
languages:
- en
- hi
- ta
- te
- ml
- ur
- bn
- mr
- gu
- or
- pa
language_creators:
- other
annotations_creators:
- no-annotation
source_datasets:
- original
size_categories:
- 1K<n<10K
licenses:
- cc-by-4-0
Dataset Card for PIB
Table of Contents
Dataset Description
- Homepage: http://preon.iiit.ac.in/~jerin/bhasha/
- Paper: https://arxiv.org/abs/2008.04860
- Point of Contact: cvit-bhasha@googlegroups.com
Dataset Summary
This dataset is the large scale sentence aligned corpus in 11 Indian languages, viz. CVIT-PIB corpus that is the largest multilingual corpus available for Indian languages.
Languages
Parallel data for following languages [en, bn, gu, hi, ml, mr, pa, or, ta, te, ur] are covered.
Citation Information
@InProceedings{cvit-pib:multilingual-corpus,
title = {Revisiting Low Resource Status of Indian Languages in Machine Translation},
authors={Jerin Philip, Shashank Siripragada, Vinay P. Namboodiri, C.V. Jawahar
},
year={2020}
}