View of Integrating Attention Attribution and Pretrained Language Models for Transparent Discriminative Learning | Journal of Computer Technology and Software