Search papers, labs, and topics across Lattice.
4
0
4
3
Transformer-based models aren't always the only answer: SVMs offer a surprisingly competitive and efficient alternative for sentiment analysis, even when contextual understanding is key.
Fine-tuning IndoBERT on Indonesian Twitter data blows away classical ML baselines in sentiment classification, suggesting transfer learning is a must-try for low-resource NLP.
Simple models still win: Logistic Regression rivals BiLSTMs with attention for Indonesian sentiment analysis, despite the latter's architectural complexity.
A BiLSTM with a custom slang dictionary rivals AutoML in classifying the sentiment and emotion of messy, real-world Indonesian e-commerce reviews.
