This dataset represents one of the largest and most comprehensive sentiment analysis collections for the Turkish language, containing over 630,000 unique samples.
It was meticulously constructed by merging, cleaning, and deduplicating several major open-source Turkish sentiment datasets to train the Keloğlan model series.
Total Unique Samples: 631,166
Training Set: 599,607
Validation Set: 31,559
Label… See the full description on the dataset page:
https://huggingface.co/datasets/engin1123/keloglan-turkish-sentiment-analysis-dataset.
