Abstract: The prevailing techniques in textual sentiment analysis include convolutional neural networks (CNNs), bidirectional long short-term memory (BiLSTM) networks, and attention mechanisms.