Skip to main content
. 2023 Oct 9;9:e1578. doi: 10.7717/peerj-cs.1578

Table 7. Comparison of training time for all models.

Bold and underlined values indicate the best result.

Model IMDB Twitter Lap14 Rest16 Hotel WaiMai SinaWeibo DouBan
ATAE-LSTM 2,662 5,185 2,855 2,940 8,185 9,952 8,136 7,893
RAM 3,855 5,915 4,065 4,560 8,915 9,083 10,775 8,624
IAN 4,638 6,672 4,738 4,560 8,672 9,662 8,408 9,758
PBAN 4,838 6,952 4,080 5,580 8,952 10,046 7,462 9,103
BiGCN 2,843 3,731 3,726 4,004 5,838 5,679 8,246 6,389
ASGCN 2,440 3,626 3,120 3,565 7,210 8,169 7,775 7,462
DualGCN 2,254 3,510 2,998 3,728 9,125 8,765 9,840 8,875
KumaGCN 2,775 3,854 4,125 4,776 8,340 8,581 9,613 8,310
PT-BERT 3,438 4,317 4,267 6,040 5,388 6,796 8,408 11,389
BERT-Pair 3,404 9,266 4,201 4,655 8,102 7,691 10,340 11,462
RoBERTa 3,542 9,105 3,989 4,287 12,501 9,657 10,000 12,475
CG-BERT 3,757 4,548 5,251 5,767 9,403 8,195 10,083 10,760