jira877832/turkish-volunteer-ner

Turkish Volunteer NER

Fine-tuned BERTurk model for extracting entities from Turkish volunteer activity announcements.

Entity types

• PLACE
• DURATION
• PARTICIPATION_TYPE
• SKILL
• TIME
• ACTIVITY_TYPE

Performance

F1 = 0.846 on a stress test set with informal language and OOV vocabulary.

Usage

from transformers import pipeline
ner = pipeline("token-classification",
               model="francky-saah/turkish-volunteer-ner",
               aggregation_strategy="simple")
result = ner("İstanbul'da hafta sonu çevre temizliği etkinliği")
print(result)

Training data

3000 annotated Turkish sentences (BIO scheme), split 2000/500/500 for train/val/test.

Citation

Part of a project on cache-aware sequential recommendation for Turkish volunteer activities.