anilguven/distilbert_tr_turkish_news

Information

This model was developed/finetuned for news classification task for the Turkish Language. This model was finetuned via news dataset. This dataset contains 7 classes: economy, magazine, sport, politics, technology, health, and events.

• LABEL_0: economy
• LABEL_1: magazine
• LABEL_2: health
• LABEL_3: politics
• LABEL_4: sports
• LABEL_5: technology
• LABEL_6: events

Model Sources

• Dataset: https://huggingface.co./datasets/anilguven/turkish_news_dataset
• Paper: peer review (Springer)
• Finetuned from model:: https://huggingface.co./dbmdz/distilbert-base-turkish-cased

Preprocessing

You must apply removing stopwords, stemming, or lemmatization process for Turkish.

Results

• Accuracy: %97.262
• F1-score: %97.263

Citation

BibTeX: Peer review process