Ptr4BERT: Automatic Semisupervised Chinese Government Message Text Classification Method Based on Transformer-Based Pointer Generator Network

Table 1

Description of datasets.

Dataset	Language	Trainset	Testset	Category	Token
HNMes	Chinese	100,000	7,000	7	53,428,396
QDMes	Chinese	40,000	7,000	7	17,283,744