NaijaSenti is an open-source sentiment and emotion corpora for four major Nigerian languages. This project was supported by lacuna-fund initiatives. Jump straight to one of the sections below, or just scroll down to find out more.
Update (05/09/2022): We are running a SemEval competition and we release more sentiment dataset from African languages including NaiJaSenti Dataset. Visit the AfriSenti SemEval page for more information : AfriSenti-SemEval Task 12
Update (02/10/2022): Send me email (shamsuddeen2004 at gmail.com) if you need more information about the dataset.
If you use this data in your work, please cite:
@misc{muhammad2022naijasenti,
title={NaijaSenti: A Nigerian Twitter Sentiment Corpus for Multilingual Sentiment Analysis},
author={Shamsuddeen Hassan Muhammad and David Ifeoluwa Adelani and Sebastian Ruder and Ibrahim Said Ahmad and Idris Abdulmumin and Bello Shehu Bello and Monojit Choudhury and Chris Chinenye Emezue and Saheed Salahudeen Abdullahi and Anuoluwapo Aremu and Alipio Jeorge and Pavel Brazdil},
year={2022},
eprint={2201.08277},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
Find the annotated dataset for Hausa, Yoruba, Igbo and Pidgin : Manually Annotated Twitter Sentiment Dataset
- Read the
NaijaSentipaper: NaijaSenti: A Nigerian Twitter Sentiment Corpus for Multilingual Sentiment Analysis - Read the
NaijaSenti Datasheetcoming soon...
Sentiment analysis is one of the most widely studied applications in NLP, but most work focuses on languages with large amounts of data. We introduce the first large-scale human-annotated Twitter sentiment dataset for the four most widely spoken languages in Nigeria—Hausa, Igbo, Nigerian-Pidgin, and Yorùbá—consisting of around 30,000 annotated tweets per language (except for Nigerian-Pidgin), including a significant fraction of code-mixed tweets. We propose text collection, filtering, processing, and labelling methods that enable us to create datasets for these low-resource languages. We evaluate a range of pre-trained models and transfer strategies on the dataset. We find that language-specific models and language-adaptive fine-tuning generally perform best. We make the datasets, trained models, sentiment lexicons, and code available to encourage sentiment analysis research in under-represented languages.
Our model is available via Hugginface Model Hub here
Please, let us know if you use NaijaSenti in your papers.
If you want to report a problem or suggest an enhancement we'd love for you to open an issue at this github repository because then we can get right on it. But you can also contact us by email (hausanlp AT gmail DOT com) or on twitter.
- 2022-01-21: Released NaijaSenti v1.0.0
Shield:
This work is licensed under a Creative Commons Attribution 4.0 International License.
