A Dual-Channel Framework for Sarcasm Recognition by Detecting Sentiment Conflict

Anonymous

Abstract: Sarcasm employs ambivalence, where one says something positive but actually means negative, vice versa. The essence of sarcasm, which is also a sufficient and necessary condition, is the conflict between literal and implied sentiments. However, it is difficult to recognize the sentiment conflict because more than one mixed or even implicit sentiments coexist in one text. As a result, the recognition of sophisticated and obscure sentiment brings in a great challenge to sarcasm detection. In this paper, we propose a dual-channel framework by modeling both literal and implied sentiment separately. Based on the flexible dual-channel framework, we design Dual-Channel Net (DC-Net) to recognize the sentiment conflict. Experiments on political debates (i.e. IAC-V1 and IAC-V2) and Twitter datasets show that our proposed DC-Net achieves state-of-the-art performance on sarcasm recognition.