uk-ner Open-Source Ukrainian Named Entity Recognition Model - Accurately Identify Names of People, Places, and Organizations in Text

Uk Ner

Developed by ukr-models

A Ukrainian named entity recognition model fine-tuned based on XLM-RoBERTa-Uk, capable of identifying person names, locations, and organization names in text.

Sequence Labeling

Transformers

OtherOpen Source License:MIT #Ukrainian NER #Named Entity Recognition #Multilingual Support

Downloads 84

Release Time : 4/7/2022

Model Overview

This model is specifically designed for named entity recognition tasks in Ukrainian texts, accurately annotating entities such as person names, locations, and organization names.

Model Features

Ukrainian Language Optimization

Fine-tuned based on XLM-RoBERTa-Uk, specifically optimized for Ukrainian texts.

Multi-category Recognition

Capable of identifying and annotating various entity types such as person names (B-PER/I-PER), locations (B-LOC/I-LOC), and organizations (B-ORG/I-ORG).

Word-level Prediction

Provides word-level rather than token-level prediction results, enhancing readability.

Model Capabilities

Ukrainian Text Analysis

Named Entity Recognition

Person Name Recognition

Location Recognition

Organization Recognition

Use Cases

Text Analysis

Ukrainian Text Entity Extraction

Extract person names, locations, and organization information from Ukrainian news, articles, and other texts.

Accurately identifies various named entities in Ukrainian texts.

Information Processing

Ukrainian Document Automation Processing

Automatically process Ukrainian documents to extract key entity information.

Improves document processing efficiency and reduces manual annotation efforts.

🚀 Ukrainian NER Model

This project is a fine - tuned Ukrainian Named Entity Recognition (NER) model that can accurately identify entities such as persons, locations, and organizations in Ukrainian text.

🚀 Quick Start

This model is a fine - tuned version of the [XLM - RoBERTa - Uk](https://huggingface.co/ukr - models/xlm - roberta - base - uk) model on the [synthetic NER dataset](https://huggingface.co/datasets/ukr - models/Ukr - Synth) with B - PER, I - PER, B - LOC, I - LOC, B - ORG, I - ORG tags.

💻 Usage Examples

Basic Usage

You can use the Huggingface pipeline to get tokens with labels:

from transformers import pipeline, AutoTokenizer, AutoModelForTokenClassification

tokenizer = AutoTokenizer.from_pretrained('ukr-models/uk-ner')
model = AutoModelForTokenClassification.from_pretrained('ukr-models/uk-ner')

ner = pipeline('ner', model=model, tokenizer=tokenizer)
ner("Могила Тараса Шевченка — місце поховання видатного українського поета Тараса Шевченка в місті Канів (Черкаська область) на Чернечій горі, над яким із 1939 року височіє бронзовий пам'ятник роботи скульптора Матвія Манізера.")

Advanced Usage

If you want to get predictions split by words instead of tokens, you can use the following approach. Download the script get_predictions.py from the repository, which uses the package tokenize_uk for splitting:

from transformers import AutoTokenizer, AutoModelForTokenClassification
from get_predictions import get_word_predictions

tokenizer = AutoTokenizer.from_pretrained('ukr-models/uk-ner')
model = AutoModelForTokenClassification.from_pretrained('ukr-models/uk-ner')

get_word_predictions(model, tokenizer, ["Могила Тараса Шевченка — місце поховання видатного українського поета Тараса Шевченка в місті Канів (Черкаська область) на Чернечій горі, над яким із 1939 року височіє бронзовий пам'ятник роботи скульптора Матвія Манізера."])

📄 License

This project is licensed under the MIT license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご