BNS-Net: A Dual-Channel Sarcasm Detection Method Considering Behavior-Level and Sentence-Level Conflicts

Sarcasm detection aims to determine whether a given utterance is sarcastic. Over the past decade, sarcasm detection has evolved from classical pattern recognition to deep learning approaches, where features such as user profile, punctuation and sentiment words have been commonly employed for sarcasm detection. However, in real-life sarcastic expressions, behaviors without explicit sentimental cues often serve as carriers of implicit sentimental meanings. Motivated by this observation, we propose a dual-channel sarcasm detection model named BNS-Net. The model considers behavior and sentence conflicts in two channels. The Behavior-level Conflict Channel reconstructs the text based on core words while leveraging the conflict attention mechanism we proposed to highlight conflict information. The Sentence-level Conflict Channel introduces external sentiment knowledge to segment the text into explicit and implicit sentences, capturing conflicts between them. To validate the effectiveness of BNS-Net, several comparative and ablation experiments are conducted. The experimental results demonstrate that the BNS-Net achieves the state-of-the-art performance.