Impact analysis of emotion in figurative language

摘要: Due to the implicit traits embedded in tweets, handling figurative languages appear as most trending topics computational linguistics. While recognition of a single language is hard capture, differentiating several at once challenging task. To achieve this purpose, we employ set emotion-based features order individuate between humor, irony, sarcasm, satire and true. We use eight basic emotions excerpted from EmoLex supplement with tweets polarity. apply these two datasets: balanced dataset (collected using hashtag-based approach) class-imbalanced streaming tweets). As result, model not only outperform word-based baseline but also handle both datasets multi-figurative detection.