metadata

license: openrail

European Languages Multipurpose Dataset for NER

Introduction

This is a multipurpose dataset that includes names, proper nouns, and acronyms from various European languages. It is designed to be particularly useful for Named Entity Recognition (NER) tasks.

Language Composition

The dataset predominantly features data in English, followed by Spanish, French, and Romanian.

Objective

The primary aim of this dataset is to further fine-tune base BERT or RoBERTa models for tasks related to Named Entity Recognition (NER).

Dataset Content

Names: Includes first names, last names, and full names.
Proper Nouns: Includes names of places, organizations, etc.
Acronyms: Includes acronyms commonly used in the featured languages.

Usage

This dataset can be used to:

Fine-tune BERT or RoBERTa models for NER tasks
Evaluate the effectiveness of different fine-tuning techniques on NER
Conduct multilingual NER research

Contributing

If you have suggestions for improving or expanding the dataset, please feel free to open an issue or submit a pull request.