Skip to main content
. 2022 Oct 22;53(11):14249–14268. doi: 10.1007/s10489-022-04221-9

Table 2.

Details of the used experimental datasets

Dataset Type of data #Classes Exceeding ratio Vocabulary size #Training set #Testing set
DATASET-1 BINARY 2 0.471 16449 6911 1821
DATASET-2 BINARY 2 0.495 39515 15000 5000
DATASET-3 MULTICLASS 8 0.297 13737 4937 2175
DATASET-4 MULTICLASS 3 0.563 16230 10000 3000
DATASET-5 MULTICLASS 5 0.553 24742 6000 1500

Note: Exceeding ratio = number of samples with a length greater than the average length/number of all training samples