gbert-legal-ner
| Property | Value |
|---|---|
| Parameter Count | 109M |
| Framework | PyTorch, Transformers |
| Language | German |
| Citation | ICAART 2023 |
What is gbert-legal-ner?
gbert-legal-ner is a specialized German BERT model designed for legal named entity recognition (NER). Developed by PaDaS-Lab, this model can identify and classify 18 different types of legal entities in German texts, including persons, judges, lawyers, organizations, laws, and court decisions.
Implementation Details
The model is implemented using the Transformers library and PyTorch framework. It can be easily integrated into existing NLP pipelines using the Hugging Face transformers library. The model processes German legal texts and identifies entities using a token classification approach.
- Built on BERT architecture with 109M parameters
- Supports 18 distinct legal entity classes
- Optimized for German legal domain
- Includes safetensors implementation
Core Capabilities
- Recognition of legal professionals (judges, lawyers)
- Identification of legal documents and references
- Detection of organizational entities
- Classification of geographical and jurisdictional entities
- Recognition of legal norms and regulations
Frequently Asked Questions
Q: What makes this model unique?
This model is specifically trained for the German legal domain with an extensive classification system covering 18 different entity types, making it highly specialized for legal document processing and analysis.
Q: What are the recommended use cases?
The model is ideal for legal document analysis, automated legal research, compliance checking, and legal information extraction from German texts. It's particularly useful for law firms, legal research institutions, and legal tech applications.
