A New Method of Improving BERT for Text Classification.
Shaomin ZhengMeng YangPublished in: IScIDE (2) (2019)
Keyphrases
- text classification
- high accuracy
- detection method
- experimental evaluation
- objective function
- preprocessing
- pairwise
- clustering method
- high precision
- synthetic data
- mutual information
- computational cost
- multiscale
- cost function
- support vector machine svm
- significant improvement
- similarity measure
- segmentation method
- classification method
- data mining