Model
Turkish Sentiment Analysis · TSAM
Purpose
Give the model a sentence in Turkish and it scores three possible readings of its tone. The published configuration maps 0 to negative, 1 to neutral and 2 to positive. The neutral label is spelled “Notr” in the output. If you use the model to tally reviews, that third option lets you count neutral comments separately.
One model, four Turkish datasets
An independent study tested four models on four Turkish datasets. In its binary evaluation, the researchers excluded neutral labels. This model scored 0.97 F1 on Humir and 0.54 on Twt.
12,053Total Hub downloadsNov 18, 2026Hugging Face Hub APIAn earlier review classifier
I fine-tuned Turkish BERT to tell positive reviews from negative ones. The training settings are on GitHub.