Research Article
Developing an Intelligent System with Deep Learning Algorithms for Sentiment Analysis of E-Commerce Product Reviews
Table 2
The splitting of the dataset.
| Total number of reviews | Training set 80% | Validation set 10% | Testing set 20% |
| 13,057 (11,184 positive; 1,873 negative) | 9,400 | 1,045 | 2,612 |
|
|