Table 9 Embedding vs classifier performance.
From: Qwen TextCNN and BERT models for enhanced multilabel news classification in mobile apps
Embedding | Best classifier | Accuracy | Worst classifier | Accuracy |
|---|---|---|---|---|
BERT-base-uncased | LSTM/MLP | 0.309 | TextCNN | 0.289 |
Multilingual-BERT | LSTM/MLP/Linear | 0.309 | TextCNN | 0.304 |
Random-embedding | LSTM | 0.309 | TextCNN | 0.305 |
Word2Vec-Style | All | 0.309 | – | 0.309 |
