“…Deep learning approaches like Convolutional Neural Networks (CNN) (Lecun et al, 1998;Teodoro et al, 2020), Recurrent Neural Networks (RNN) (Rumelhart et al, 1986), Long Short-Term Memory Networks (LSTM) (Hochreiter and Schmidhuber, 1997), and Transformer-based architectures (Vaswani et al, 2017), including pretrained language models such as BERT (Devlin et al, 2018), RoBERTa (Liu et al, 2019), and XL-Net (Yang et al, 2019), have demonstrated stateof-the-art efficacy in a diverse range of domains . Leveraging the hierarchical structure of documents, graph neural networks (GNNs) have also been effectively proposed to assign categories to biomedical documents (Ferdowsi et al, 2023(Ferdowsi et al, , 2022(Ferdowsi et al, , 2021. Compared to deep learning models, SVM requires lower computational resources and training time and is a more efficient choice for certain applications (Sakr et al, 2016).…”