Sentiment Classification on Tokopedia Product Reviews 2025
97.6AccuracyTF-IDF + Linear SVC
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TF-IDF + Linear SVCFeature Extraction=TF-IDF2026.05 | 97.6 | 55.1 | 97.4 | |
| TF-IDF + Multinomial Naive BayesFeature Extraction=TF-IDF2026.05 | 97.5 | 32.9 | 96.3 | |
| TF-IDF + Logistic RegressionFeature Extraction=TF-IDF2026.05 | 94.36 | 51.64 | 95.75 | |
| IndoBERT (Fine-tuned)Evaluation Protocol=Fine-tuned, batch size=16, learning rate=2 × 10−5, epochs=3, maximum sequence length=256, warmup ratio=0.1, weight decay=0.012026.05 | 88.7 | 50.88 | 92.68 |