Sarcasm detection is a specific case of sentiment analysis in natural language processing (NLP). Irony text and sarcasm detection aim to recognize the sarcasm in the utterances along with inconsistency between positive words and the negative context in the sentence. Traditional deep learning models focus on automatic sarcasm detection and possess the limitation to identifying the incongruity between words. Attention mechanism in deep learning is introduced to capture incongruity between positive words and the negative context in the sarcastic sentence.
The significance of an attentive model is to focus on word-level information to understand the intention of the sentence. Deep attentive model based irony text and sarcasm detection provide the best performance and help in determining the meaningful insights of sarcasm and irony text.
• In the advancement of Artificial Intelligence (AI), misunderstanding irony and sarcasm constitute a big challenge.
• Irony and sarcasm detection expedite the understanding of the real intentions of the human behind the ironic and sarcastic language.
• In order to ameliorate the irony and sarcasm detection, the attention mechanism with the deep learning model is exploited.
• An attention model with a deep neural network supports sentimental knowledge in recognizing the explicit and implicit context incongruity of irony and sarcasm detection.
• The deep attentive model learns deeper contextual word relationships in ironic statements and amplifies performance by capturing the contribution of essential words for efficient irony and sarcastic detection.