Ner Medical Japanese

🚀 Japanese Medical Conversation NER Model

This model is a Named Entity Recognition (NER) model specifically designed for Japanese medical conversation documents. It is fine-tuned based on tsmatz/xlm-roberta-ner-japanese and optimized for information extraction and anonymization in the medical field.

🚀 Quick Start

from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline

model = AutoModelForTokenClassification.from_pretrained("Tetsuo3003/ner-medical-japanese")
tokenizer = AutoTokenizer.from_pretrained("Tetsuo3003/ner-medical-japanese")
ner_pipeline = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")

text = "金丸先生が松本市にある石川クリニックに通院しました。"
results = ner_pipeline(text)
for entity in results:
    print(f"{entity['word']} → {entity['entity_group']} (score: {entity['score']:.2f})")

✨ Features

• Named Entity Extraction: Extract named entities (such as names, organization names, and place names) from medical conversation documents.
• Anonymization: Perform anonymization processing on medical data.
• Data Organization: Organize information in medical-related data.
• Risk Analysis and Research Data Organization: Assist in risk analysis and research data organization.

📚 Documentation

Target Domains

• Conversation Records: Records of conversations with patients.
• Medical Notes: Medical records, home nursing records.
• Reports and Documents: Reports and documents in medical facilities.

Label List

📄 License

This project is licensed under the MIT license.