Multilingual Complex Named Entity Recognition (MultiCoNER)

Dataset Summary

MultiCoNER (version 1) is a large multilingual dataset for Named Entity Recognition that covers 3 domains (Wiki sentences, questions, and search queries) across 11 languages, as well as multilingual and code-mixing subsets. This dataset is designed to represent contemporary challenges in NER, including low-context scenarios (short and uncased text), syntactically complex entities like movie titles, and long-tail entity distributions. The 26M token dataset is compiled from public resources using techniques such as heuristic-based sentence sampling, template extraction and slotting, and machine translation.

See the AWS Open Data Registry entry for MultiCoNER for more information.

Labels

• PER: Person, i.e. names of people
• LOC: Location, i.e. locations/physical facilities
• CORP: Corporation, i.e. corporations/businesses
• GRP: Groups, i.e. all other groups
• PROD: Product, i.e. consumer products
• CW: Creative Work, i.e. movies/songs/book titles

Dataset Structure

The dataset follows the IOB format of CoNLL. In particular, it uses the following label to ID mapping:

{
    "O": 0,
    "B-PER": 1,
    "I-PER": 2,
    "B-LOC": 3,
    "I-LOC": 4,
    "B-CORP": 5,
    "I-CORP": 6,
    "B-GRP": 7,
    "I-GRP": 8,
    "B-PROD": 9,
    "I-PROD": 10,
    "B-CW": 11,
    "I-CW": 12,
}

Languages

The MultiCoNER dataset consists of the following languages: Bangla, German, English, Spanish, Farsi, Hindi, Korean, Dutch, Russian, Turkish and Chinese.

Usage

from datasets import load_dataset

dataset = load_dataset('tomaarsen/MultiCoNER', 'multi')

License

CC BY 4.0

Citation

@misc{malmasi2022multiconer,
    title={MultiCoNER: A Large-scale Multilingual dataset for Complex Named Entity Recognition}, 
    author={Shervin Malmasi and Anjie Fang and Besnik Fetahu and Sudipta Kar and Oleg Rokhlenko},
    year={2022},
    eprint={2208.14536},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}