bert-base-uncased-city-country-ner
| Property | Value |
|---|---|
| Model Type | Named Entity Recognition (NER) |
| Base Architecture | BERT-base-uncased |
| Author | ml6team |
| Model URL | https://huggingface.co/ml6team/bert-base-uncased-city-country-ner |
What is bert-base-uncased-city-country-ner?
This specialized NER model is fine-tuned on a custom dataset to specifically identify city and country names within text. Built upon the BERT-base-uncased architecture, it provides accurate geographical entity recognition through a three-tag classification system: OTHER, CITY, and COUNTRY.
Implementation Details
The model leverages weakly supervised learning on the Ultra-Fine Entity Typing dataset, enhanced with city and country information. Additional preprocessing steps were implemented to reduce false labels and improve accuracy. The implementation uses the Transformers library and can be easily integrated into existing NLP pipelines.
- Built on BERT-base-uncased architecture
- Custom dataset with weak supervision
- Three-tag classification system
- Optimized preprocessing for label accuracy
Core Capabilities
- Accurate identification of city names in text
- Reliable country name detection
- Simple integration with Transformers pipeline
- Support for case-insensitive text processing
Frequently Asked Questions
Q: What makes this model unique?
This model specializes in geographical entity recognition, specifically optimized for identifying cities and countries. Its focused scope and custom dataset training make it particularly effective for geographical text analysis.
Q: What are the recommended use cases?
The model is ideal for applications requiring geographical entity extraction, such as travel content analysis, address processing, location-based services, and geographical data mining from unstructured text.
