Published March 17, 2021 | Version V1.0
Dataset Open

COMPRISE_Data11_M-NER_V1.0

Authors/Creators

  • 1. Universität des Saarlandes

Description

A named entity recognition (NER) dataset (a subset of the News dataset). It has four entity types: Personal names, Locations, Organizations, and Dates. The dataset supports ten African languages: Amharic, Hausa, Igbo, Kinyarwanda, Luganda, Luo, Nigerian-Pidgin, Swahili, Wolof, and Yoruba.

Files

COMPRISE_Data11_M-NER_V1.0.zip

Files (2.2 MB)

Name Size Download all
md5:0caa4bf3ce097f0a1d1975e1538391d4
2.2 MB Preview Download

Additional details

Funding

European Commission
COMPRISE - Cost-effective, Multilingual, Privacy-driven voice-enabled Services 825081