metadata
tags:
- textual-entailment
language:
- ind
This dataset is designed for Natural Language Inference NLP task. It is designed to provide a challenging test-bed for Indonesian NLI by explicitly incorporating various linguistic phenomena such as numerical reasoning, structural changes, idioms, or temporal and spatial reasoning.
Dataset Usage
Run pip install nusacrowd before loading the dataset through HuggingFace's load_dataset.
Citation
title = "{I}ndo{NLI}: A Natural Language Inference Dataset for {I}ndonesian",
author = "Mahendra, Rahmad and Aji, Alham Fikri and Louvan, Samuel and Rahman, Fahrurrozi and Vania, Clara",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.821",
pages = "10511--10527",
}
License
Creative Common Attribution Share-Alike 4.0 International
Homepage
NusaCatalogue
For easy indexing and metadata: https://indonlp.github.io/nusa-catalogue